Skip to main content

Full text of "Weather Elements Fouth Edition"

See other formats





(< OU_1 60468 >m 


Call No. J S7-S" | &*><** Accession No. 



This book shofd> returne'd'on or bfefore\he date last marked belgw. 





Former Senior Metcoioloirist, U. S \Vealhci Bmeau 
Formet Associate* Professor ol Meteoiolo^v 
sity o( Nrhriska 



Professor of Orography <iiul Meteoiology 
Oklalionia State Um\eisi(v 


En^leicood Cliffs, N.J. 

COPYRIGHT, 1937, 1942, 1948, 1957, BY 



CATALOG CARD No.: 57-5047 

Find printing /February, 1957 

Second printing I. January, 1958 




The science of meteorology has developed very rapidly during the 
last decade. Impetus received from World War II and the Korean 
conflict has stimulated research and development which indicate 
the possibility of foretelling the weather with much greater accuracy 
within the next few years. The discovery of rain-stimulating tech- 
niques, using Dry Ice and silver iodide, has opened an entirely 
new possibility of weather control and has contributed to the crea- 
tion of a new professional group, known as private wither con- 
sultants, in the United States. 

This book is intended not only to serve as a foundation for ad- 
vanced study in meteorology, but also to furnish the basic weather 
knowledge that is essential to airplane pilots, geographers, agricul- 
turists, biologists, engineers, industrialists, and many sciences in 
which weather is an important factor. It should be desirable for 
individual as well as for classroom study. Numerous footnotes and 
a selected bibliography have been included to guide those who wish 
to pursue the subject beyond the scope of this book. 

The present edition has been completely revised to incorporate 
the new developments in meteorology and to remove those older 
concepts that are no longer representative of the science. Care has 
been taken to avoid the use of technical descriptions and mathemat- 
ical formulas which are beyond the comprehension of the average 
college freshman. New illustrative materials have been used unspar- 
ingly to increase reader interest and understanding. Problems at 
the end of each chapter are provided to supplement the text. In most 
cases, answers cannot be found directly in the text but require 
comprehension and application of the .text materials. Experience in 
teaching elementary meteorology for several years has shown that 
problems of this nature benefit the more capable students especially. 

I am indebted to the U. S. Navy for much basic training and 
practical experience in meteorology. It was my privilege, while on 


active military duty, to study under such leading meteorologists as 
Professors J. Bjerknes, Jorgen Holmboe, Morris Neiburger, and 
H. U. Sverdrup. While serving as aerological officer in the Pacific 
theater during World War II, I made many typhoon reconnaissance 
flights, as observer, from Clark Field, Philippine Islands. This first- 
hand experience with the dynamics of the atmosphere can hardly 
be equalled by any other type of weather training. 

I am especially indebted to Mr. W. E. Maughan, former Section 
Director, U. S. Weather Bureau, Oklahoma City, and to Professor 
W. E. Hardy, former Head of the Department of Meteorology, Okla- 
homa State University, for reading the manuscript. Any errors 
remaining, however, are due to my own inadvertence. My thanks 
are extended to the many contributors of illustrative materials 
( credit is given individually after the figures ) and to my colleagues 
for their patience, consideration, and encouragement. 

Oklahoma State University 
Stillwater, Oklahoma 


This book aims to present, concisely and systematically, an intro- 
duction to the science of meteorology in its present stage of develop- 
ment. My primary purpose is to set forth the facts and principles 
concerning the behavior and responses of the atmosphere in such a 
way as to enable the reader to acquire an elementary understanding 
of the physical processes underlying observed weather phenomena, 
An important secondary object is to present that general body of 
information about the weather and the present state of our knowl- 
edge concerning it which, it is believed, every intelligent person 
should possess in relation to this most important element of his 

Attention is given to the instruments and methods used in observ- 
ing and measuring atmospheric conditions, to the complex effects of 
solar radiation, and to the interrelations of the various weather ele- 
ments. Other subjects treated are the general circulation of the 
atmosphere and its modifications, the basis of weather forecasting, 
the general geographic distribution of the weather elements, and 
some of the relations of weather and climate to man and his varied 
life. A brief account is given of the electrical and optical phenom- 
ena of the atmosphere, and of the organization and activities of the 
United States Weather Bureau. Some knowledge of such matters is 
a necessary foundation for the study of geography, agriculture, and 
ecology, and of aeronautics, hydrology, and other branches of engi- 
neering. In addition, such knowledge is useful in a great variety of 
professions and occupations, notably in medicine, law, and business. 
It is hoped, therefore, that the book will prove of value to persons of 
widely varying interests. 

The discussions are necessarily brief and incomplete, and much 
interesting and valuable material has been omitted, but it is hoped 
that the most important phases of the subject have been treated in 
such a way as to arouse an abiding interest and lead to further read- 
ing. Meteorology is a growing science and at the present time is 



undergoing a rather rapid development and transition. An effort 
is made to present the modern aspects of the subject and to indicate 
the lines along which research is being conducted and progress being 

The author of a book on elementary meteorology is inevitably in- 
debted to the pioneers and leaders in the science of the weather in 
past generations, and also to a great number of contemporary stu- 
dents and investigators. No one can write such a book without be- 
coming aware of meteorology's debt to Sir Napier Shaw and Dr. 
W. J. Humphreys, among present-day scientists. Dr. H. C. Willett 
and Mr. Jerome Namias have recently contributed notably to a 
knowledge of air masses and air-mass analysis. Nearly all the writers 
listed in the bibliography have provided facts or ideas which have 
been drawn upon in the preparation of this text. I regret that it 
has not been possible to identify and acknowledge the original 
source in each case. For general scope and for the general method 
of treatment of some of the fundamentals the author is conscious of 
the influence of the older textbooks, especially those by W. M. Davis 
and W. I. Milham. 

To Dr. Nels A. Bengtson I am deeply grateful, not only for his 
critical reading of the entire manuscript, but also for his sympathetic 
interest in the book and his help in many of the details of its prepara- 
tion. I am indebted to the Chief of the United States Weather Bu- 
reau for permission to publish this work, and to the scientific staff at 
the Central Office of the Weather Bureau for valuable criticisms and 
suggestions. My thanks are extended to Mr. D. Keith Kinsey, who 
prepared nearly all the drawings, to Mr. H. Floreen for his cloud 
photographs, and to Professor J. C. Jensen, Mr. Otto Wiederanders, 
the United States Weather Bureau, and Julien P. Friez and Sons, for 
other photographic illustrations. The Taylor Instrument Companies 
have kindly given permission for the use of some material originally 
published in Tycos-Rochester (now Taylor-Rochester), and the 
Denoyer-Geppert Company has permitted the use of its base maps. 
My wife has given valued assistance and encouragement and has 
helped in reading the proofs. 



I. The Atmosphere 1 

^-^introduction, 1. Composition of the Atmosphere, 5. Properties of 
the Atmosphere, 8. The Elements of Weather, 11. Problems, 13. 

II. Observing Temperature, Pressure, and Wind 14 

Temperature Observations, 14. Pressure Observations, 24. Wind 
Observations, 32. Problems, 41. 

III. Observing Moisture, Sunshine, Visibility, and Upper Air 
Conditions 42 

Humidity Observations, 42. Evaporation Observations, 51. Cloud 
Observations, 55. Precipitation Observations, 67. Sunshine Ob- 
servations, 71. Observations of Visibility and Ceiling, 73. Upper 
Air Observations, 75. Problems, 80. 

IV. Solar Radiation 82 

Radiant Energy, 82. Incoming Solar Radiation, 84. Amount of 
Insolation Received at a Fixed Location, 87. Direct Eifects of 
Solar Radiation, 91. Conduction, 96. Convection, 98. Prob- 
lems, 99. 

V. Adiabatic Processes and Stability 100 

Adiabatic Temperature Changes, 100. Lapse Rates, 106. Stability 
and Instability, 109. Atmospheric Layers, 116. Problems, 120. 

VI. Condensation of Water in the Atmosphere 122 

Condensation on Solid Surfaces, 122. Condensation Above the 
Earth's Surface, 126. Clouds and Precipitation, 131. Forms of 
Precipitation, 135. Artificial Rain Stimulation, 139. Problems, 142. 

VII. Interrelations of Temperature, Pressure, and Wind 143 

Pressure Gradients, 143. Gradient Winds and Surface Winds, 
. Winds Due to Local Temperature Differences, 150. Mon- 
s, 153. Polar-Equatorial Air Movements, 154. Problems, 155. 

VIII. The General Circulation 157 

Yearly Averages of Pressure, 157. January and July Averages of 
Pressure and Winds, 162. Winds Aloft, 168. Latitudinal Inter- 
change of Air, 170. Problems, 173. 

IX. Air Masses and Fronts " 174 

Air Masses, 174. Characteristics of North American Air Masses, 
178. Formation and Characteristics of Fronts, 183. Problems, 189. 



X. The Secondary Circulation 191 

Weather Maps, 191. Low Pressure Centers, 193. High Pressure 
Centers, 200. Nature and Origin of Highs and Lows, 201. At- 
mospheric Circulation in the Tropics, 208. Problems, 211. 

XI. Special Storms and Lesser Atmospheric Disturbances 212 

Tropical Cyclones, 212. Thunderstorms, 221. Tornadoes, 230. 
Some Special Winds, 238. Problems, 241. 

XII. Weather Analysis and Forecasting 243 

Analysis of Synoptic Charts, 243. Upper-Air Analysis, 247. Fore- 
casting the Weather, 259. Local Forecasts and Weather Lore, 
273. Problems, 277. 

XIII. Aviation and the Weather 278 

Airways Service, 278. Flying Weather, 281. Problems, 291. 

XIV. Electrical and Optical Phenomena 292 

Electrical Phenomena, 292. Optical Phenomena, 295. Problems, 

XV. Climate 302 

Climatic Elements, 302. Distribution of Temperature, 306. Gen- 
eral Distribution of Precipitation, 315. Zonation of Climates, 322. 
Climate as Related to the Physical Features of the Earth's Sur- 
face, 325. Climatic Classification, 328. Cyclical Changes of 
Weather and Climate, 330. Theories of Climatic Changes, 336. 
Problems, 339. 

XVI. World Weather Relations and Climatic Influences 340 

Variability of the Weather, 340. Persistence of the Weather, 343. 
Weather Correlations, 347. The Oceans and the Weather, 348. 
The Sun and the Weather, 351. Seasonal Forecasting, 352. 
Weather and Health, 354. Climate and Culture, 359. Agricultural 
Meteorology, 362. Problems, 365. 

XVII. Weather Services of the United States 366 

Brief History of Meteorology, 366. Development of a Weather 
Service, 367. Present Organization of the Weather Bureau, 369. 
Activities of the Weather Bureau, 370. Military and Private 
Weather Services, 375. Prospect for the Weather Services, 377. 

Appendix I. Bibliography 379 

Appendix II. Conversion Factors and Tables 385 

Appendix III. Mean Monthly and Annual Temperatures and 

Precipitation (Tables) 390 

Index 397 


1. The Earth's Atmosphere and Atmospheric Phenomena .... 3 

2. Composition of Dry Air by Volume 6 

3. Relation of Volume to Pressure on a Mass of Gas 10 

4. Thermometer Scales Compared 17 

5. Maximum and Minimum Thermometers with Townsend 
Support 18 

6. Thermograph 19 

7. Instrument Shelter with Door Open 20 

8. Thermograph Record 22 

9. Typical Curves of Annual March of Temperature 23 

10. A Simple Mercurial Barometer 24 

11. Mercurial Barometers Mounted in Case 25 

12. Barometer Scales Compared 27 

13a. Schematic Drawing of an Aneroid Barometer 28 

13b. Navy Type Aneroid Barometer 28 

14. Barograph 29 

15. Wind Instruments 33 

16. Wind Directions 34 

17. Beaufort Wind Scale 35 

18. Robinson 3-cup Anemometer 36 

19. Aerovane 37 

20. Anemoscope Record 38 

21. Wind Rose for New York City 40 

22. Whirled Psychrometer 47 

23. Weekly Hygrograph Record 52 

24. Class A Evaporation Station 54 

25. Generalized Vertical Arrangement of Cloud Types 58 

26 through 40. Cloud Types 59-66 

41. Rain Gage 68 

42. Fergusson Weighing Rain Gage 69 

43. Electrical Sunshine Recorder 72 

44. Measuring Ceiling Height 74 

45. Radiosonde in Flight 77 

46. Electromagnetic Spectrum 83 



47. Pyrheliorneter 86 

48. Earth's Orbit About the Sun 89 

49. Effect of Angle of Incidence 90 

50. Diagrammatic Representation of the Distribution of Solar 
Radiation 93 

51. Transfer of Heat by Convection 98 

52. Convectional Circulation 99 

53. Adiabatic Changes of Temperature 102 

54. Mechanics of Equivalent Potential Temperature 104 

55. The Adiabatic Chart 105 

56. Examples of Lapse Rates 108 

57. Stable Atmospheric Condition 110 

58. Unstable Atmospheric Condition Ill 

59. Conditional Instability 113 

60. Instability of Lifting 114 

61. Layers of the Atmosphere 117 

62. Orchard Heaters 125 

63. Wind Machine for Fighting Freezes 126 

64. Penetrative Convective Cumuli 133 

65. Snow Crystals 136 

66. Typical Large Hail Stones 138 

67. Glaze Ice 140 

68. Cloud Seeding Generator 141 

69. Isobars and Pressure Gradient 144 

70. Vertical Cross-section of Isobaric Surfaces 145 

71. Effect of the Coriolis Force 146 

72. Three Forces Affecting the Wind 148 

73. Surface Wind in Relation to Isobars 149 

74. Sea Breeze 150 

75. Land Breeze 151 

76. Mean Annual Sea Level Pressure 158 

77. Pressure and Wind Belts 160 

78. Schematic Representation of the Cells of Atmospheric Cir- 
culation 161 

79. Mean January Sea Level Pressures 163 

80. Mean July Sea Level Pressures 164 

81. East- West Transport of Air 168 

82. Three-dimensional Cold Front 184 

83. Warm Front 185 

84. Cold Front 187 

85. Station Model 192 

86. Weather Map 194 

87. Weather Map 195 


88. Weather Map 196 

89. Barometric Depression 197 

90. Typical Paths of Cyclones 199 

91. Anticyclone 200 

92. Typical Paths of Anticyclones 201 

93. Stages of a Wave Cyclone 204 

94. Frontal Occlusions 206 

95. Wedge of Cold Air 208 

96. Pressure Profile of the Tropics 209 

97. Streamline Chart 209 

98. Eye of a Typhoon 213 

99. Paths of Tropical Hurricanes 216 

100. Tropical Hurricane 219 

101. Thunderstorm 222 

102. Number of Days with Thunderstorms 226 

103. Direct Lightning Discharge 228 

104. Ball Lightning 230 

105. Tornado at Gothenburg, Nebraska 231, 232 

106. Tornado at Rockwell, Texas 233 

107. Small Waterspout 237 

108. Foehn Wind 239 

109. Dust Storm 241 

110. Atmospheric Cross-section 249 

111. Rossby Diagram 251 

112. Pressure Contour Chart, 500 Mb 254 

113. Isentropic Chart 256 

114. Pressure Contour Chart, 700 Mb 258 

115. Weather Map 263 

116. Weather Map 264 

117. Weather Map 265 

118. Percentage of Time Rain Occurred 274 

119. Rime and Glaze on Wing of Aircraft 283 

120. Icing in Frontal Conditions 284 

121. Carburetor Icing 286 

122. GCA Radarscope 287 

123. Eddy Turbulence in Mountains 289 

124. Aurora Band Just After Sunset 294 

125. Refraction of Light 295 

126. Atmospheric Refraction 296 

127. Mirage in Death Valley 297 

128. Path of Light of Primary Rainbow 298 

129. Halo Phenomena 299 

130. Halo of 22 300 


131. Mean Annual Sea Level Temperatures 308 

132. Mean January Sea Level Temperatures 310 

133. Mean July Sea Level Temperatures 311 

134. Mean Annual Range of Temperature 313 

135. January Mean Daily Range of Temperature 314 

136. Mean Annual Precipitation 316 

137. Types of Rainfall Distribution 318 

138. Average Number of Days with Rain 320 

139. Climates of North America 329 

140. Iowa Precipitation 332 

141. Secular Trends of Temperature 334 

142. Frequency Polygon of Precipitation 342 

143. Frequency Polygon for Growing Season 343 

144. Departure of Mean Monthly Temperatures from Normal . . . 345 

145. Facsimile 376 





Weather is a subject of universal interest. Man has been concerned 
about its vagaries for countless ages. He has observed the fury of 
mighty storms, the freshness of gently falling raindrops, the burn- 
ing sun over a drought-stricken land, and the refreshing breezes of 
a spring afternoon. These things caused him to wonder about the 
powers that controlled the elements, which could change a pleasant 
summer afternoon into a frightful display of wind, fire, hail, or flood. 
Yet, it was only recently that he began to understand the secrets of 
nature that collectively comprise or cause our weather. 

The earth, as it turns on its axis and follows its elliptical orbit 
about the sun, is forever encompassed by a gaseous shell or envelope, 
called the atmosphere. This encircling air is an integral and essential 
part of the earth. We frequently remain unaware of its presence be- 
cause of its invisibility; but we feel its force when it moves rapidly 
past us, and we know that "when the trees bow down their heads, 
the wind is passing by." It is an ever-moving air, not fixed like land, 
nor confined in basins and channels like water, but ever-present and 
all-prevailing about the face of the earth. 

Air is the medium in which we live and move, the breath of life 
to man, animals, and plants. Without it, this would be a dead and 
barren world. We can live for many days without food, for a few 
days without water, but for only a few minutes without air. The food 
that we eat and the clothing that we wear are composed in large 
part of elements obtained from the air. The water that falls on our 
fields is carried to them by the air. Air supports combustion and 
transmits sound. In addition to these primary and indispensable func- 
tions, air is made to serve us in many practical ways. It is a source 


of power to sailing ships and windmills. When compressed, it actu- 
ates the brakes of moving vehicles and the trip hammers of pneu- 
matic tools; it serves us in many other ways, and finally it furnishes 
a support and a highway for vast machines, "as the heavens fill with 
commerce." It is with the behavior of this intangible, colorless, and 
odorless but all-important atmosphere that this book deals. 

Extent of the atmosphere. The atmosphere extends to a height 
of several hundred miles above the earth's surface (Fig. 1). The ef- 
fective upper limit of the atmosphere is reached at a much lower 
altitude when the air becomes extremely rarefied. Although it seems 
light, there are approximately 11,850,000,000,000,000,000 pounds of 
air weighing down upon the earth at all times. Being compressible, 
the air is more dense near the surface of the earth than at higher 
altitudes because of pressure from the upper layers on those below. 
It becomes so rare at only 15,000 feet (5 km) elevation that supple- 
mentary oxygen must be provided for air travelers; and at 18,000 
feet (6 km), half of the mass of the atmosphere lies below. 

Very little is known of the outer atmosphere, although knowledge 
of this space is continually growing. Twilight colors give evidence 
of considerable air at a height of 40 miles (64 km); meteors and the 
aurorae give evidence of gases at much higher levels, but it is cal- 
culated that there is no appreciable atmospheric pressure above 50 
miles (80 km). 

Rocket flights have reached elevations up to 250 miles (400 km) 
and have supplied valuable but limited information about these high 
elevations. Recognition of the influence of high layers of air on solar 
radiation and radio receptions, together with an increasing interest 
in the possibilities of space travel, have caused scientists to try every 
known means of examining outer space. It is possible, also, that the 
air at extremely high altitudes may affect the surface weather con- 
ditions, either directly or indirectly. In general, however, weather is 
a low-level phenomenon. Clouds frequently are found at elevations 
of 30,000 feet (9.1 km), but rarely reach 40,000 feet (12.2 km) except 
in some thunderstorms, where they may reach more than 80,000 
feet (25 km). The study of meteorology is therefore restricted mainly 
to a study of a thin boundary layer of the atmosphere about 15 miles 
(24 km) in thickness. 

Meteorology defined. Meteorology is the science of the atmos- 
phere and its phenomena those phenomena which we call, collec- 


250 ^ 










- 50 






-I +1898* 

(Woe Corporol-250 miles) 

(114 i/iUs) 

bin, ill!.. 






NACREOUS CLOUDS' / (72,395 feet) 



(Navy jet Skyrocket-2 -79,000 feet) I 















MT. EVEREST; 29,002 FT. 


Fig. 1. The Earth's Atmosphere and Atmospheric Phenomena. By permission of the 
NEW YORK TIMES and Mr. James Lewicki. 


tively, the weather. Because of their infinite variety and their in- 
timate relation to all of our activities, the phenomena of the weather 
are subjects of never-ending interest. They are not only of interest, 
but are also of great importance, since weather is one of the chief 
elements in man's environment, far-reaching in its influence and 
affecting all phases of his life. One of the reasons for the daily inter- 
est in the weather in regions outside of the tropics is that it is new 
every morning. It is never stable for long but always in a state of 
becoming something different. In this it is typical of all nature; but 
since weather changes are more rapid and noticeable than most 
other natural changes, "changeable as the weather" has become a 
time-worn simile. 

Meteorology combines physics and geography. It not only applies 
the principles of physics to the behavior of the air treated as a mix- 
ture of gases, but it considers the whole atmosphere and its move- 
ments as they are affected by such geographic factors as latitude, 
topography, altitude, and distribution of land, water, and moun- 
tains. The geography of the globe is an essential factor in its weather. 
Insofar as meteorology deals with the physics of the air, it is a branch 
of physics; insofar as it is descriptive and explanatory of the environ- 
ment of man, as affecting his modes of life and his ways of making a 
living, it is a branch of ecology and geography. In combining the two 
to account for actual weather and climate, it is something different 
from either: it is a separate branch of science. Moreover, the facts 
and data accumulated in the study of weather and climate are ca- 
pable of infinite application to the life of man. They are of impor- 
tance in the study of history, geology, and biology. They are used 
directly and daily by the farmer and the engineer, by the physician, 
the lawyer, and the businessman. 

The science of meteorology has advanced by the following steps: 

1. The invention of instruments and methods for determining the 
condition of the air. 

2. The use of these in the systematic accumulation of observa- 
tional data. 

3. The classification and organization of the accumulated data 
for the purpose of discovering and describing the condition of the 

4. The development of physical theories to interpret and co- 
ordinate atmospheric processes. 


5. The application to useful purposes of the knowledge thus ac- 

Some of the Greek philosophers, notably Hippocrates, Aristotle, 
and Theophrastus, approached the study of weather and climate in 
a scientific spirit and made progress in interpreting the phenomena 
of the atmosphere rather remarkable progress, in view of the lim- 
ited knowledge of physics and chemistry existing in their day. We 
derive the name "meteorology" from Aristotle's treatise, Meteoro- 
logica. This treatise included a discussion of much that we now call 
astronomy, physical geography, and geology, but about one-third of 
it was devoted to atmospheric phenomena. 

During the Middle Ages, weather events were given many irra- 
tional and mystical interpretations. The weather was a matter of 
signs and portents, often assumed to be related to human conduct 
as warning, punishment, or reward. Some of the mystery remains in 
many minds, and superstition still survives when men discuss the 
weather. And yet, the elementary facts about atmospheric phenom- 
ena and their causes are simple and easily acquired, and they are nec- 
essary to an intelligent appreciation of our daily life. It is true, on the 
other hand, that much is yet unknown about the behavior of the air. 
The atmosphere is so vast as to preclude, perhaps forever, the pos- 
sibility of a complete analysis of its forces and activities. Therein lies 
an opportunity and a challenge for additional accumulation of facts 
and for further investigation and research. 

Composition of the Atmosphere 

Upon examination, the atmosphere is found to be a complex sys- 
tem, not a simple chemical element nor even a compound, but a rela- 
tively stable mixture of a number of gases. First, there are several 
chemical elements which remain permanently in gaseous form under 
all natural conditions. Second, gaseous water, known as water vapor, 
is a variable part of this mixture. Under certain conditions, liquid 
and solid forms of water also occur in the air, but these are not in- 
cluded in the definition of air. Finally, the air always contains, but 
not as essential ingredients, a great number of solid particles of vari- 
ous natures, known collectively as dust. 

Permanent gases. The two permanent gases that make up 99 per 
cent of the volume of the air, after the water vapor and the dust 


particles have been removed, are the chemical elements nitrogen and 
oxygen. These two elements, in combination with others, also make 
up a large portion of all living matter and of the earth's crust. Nitro- 
gen forms about 78 per cent of the total volume of dry air, and 
oxygen about 21 per cent (Fig. 2). Of the remaining 1 per cent, 
the greater part is argon, and only about 0.04 per cent remains, of 
which approximately 0.03 per cent is carbon dioxide and the re- 
mainder is neon, helium, krypton, hydrogen, xenon, ozone, radon, 
and other gases. 

OXYGEN- 20.99% 

NITROGEN -78.03% 

Fig. 2. Composition of Dry Air by Volume. Water vapor is also present in the air, 
but varies from near zero to 4 per cent. 

The relative percentages of the principal permanent gases remain 
remarkably constant throughout the world from the surface of the 
earth to heights of several miles. We breathe the same air every- 

The active energizing element of the air is oxygen, which com- 
bines readily with other chemical elements and is necessary to life. 
The carbon dioxide which is exhaled by animals is absorbed by 
plants, and its oxygen constituent later released to the air. This re- 
ciprocal use by plants and animals plays a part in maintaining a rela- 
tively constant ratio of these two gases. The waters of the ocean also 
exercise a control over the concentration of carbon dioxide in the air: 
when the amount of carbon dioxide increases, more is absorbed by 
the water; when it decreases, some of the gas returns to the atmos- 
phere. The other permanent gases appear to have no special natural 


functions except to increase the density of the atmosphere and to 
dilute its oxygen. But some soil bacteria take nitrogen from the air 
and make it available to plants, and man has also learned how to 
utilize atmospheric nitrogen. The rare gases neon, krypton, and 
xenon are also obtained by extraction from the air. 

Water vapor. Water vapor is contributed to the air by evapora- 
tion from water surfaces, soil, and living tissues and by combustion. 
It is an all-important constituent of the atmosphere, but, unlike the 
other gases, is quite variable in amount, ranging from a minute pro- 
portion in the air of deserts and polar regions to a maximum of ap- 
proximately 4 per cent by volume in the warm and humid tropics. 
Some water remains in the air as a gas at all temperatures; but the 
amount that may be mixed with the other gases of the air at low 
temperatures is small compared with the possible amount when the 
temperature is high. The importance of atmospheric moisture to all 
forms of life is so universally recognized that no additional emphasis 
need be given here. However, less well known is the very important 
role which it plays in the physical processes of the atmosphere. 
Water vapor in the air affects its temperature, density, and humidity 
and its heating and cooling characteristics; these will be explained 

Dust. The gases of the atmosphere maintain in suspension an 
immense number of nongaseous substances of various kinds which 
are collectively called dust. In addition to the visible dust which 
sometimes fills the air and darkens the sun in dry regions, the air 
always, or nearly always, carries small particles of organic matter, 
such as seeds, spores, and bacteria. Much more numerous, however, 
are the microscopic, inorganic particles which contribute to the for- 
mation of haze, clouds, and precipitation. Some of these are fine par- 
ticles of soil or of smoke, or salts from ocean spray, which are lifted 
and diffused by the winds and rising air currents. The dust particles 
are naturally more numerous in the lower atmosphere, but some of 
them are carried to heights of several miles. Large numbers of even 
finer particles are thrown into the air by volcanic explosions, and 
many more result from the burning of meteors in the upper air, thus 
furnishing a supply of dust to the air at great heights. 

Many of these particles are very minute, but they have two impor- 
tant effects on the weather. First, many of them are water-absorbent 
and are the nuclei on which condensation of water vapor begins. 


Second, they intercept some of the heat coming from the sun. When 
there is an unusual amount of such dust, as in a time of great vol- 
canic activity, the result may be to reduce the average temperature 
of the globe, Dust plays a part in the creation of the varied colors 
of sunrise and sunset. For three years after the violent explosion of 
the volcano, Krakatoa, in the East Indies, in 1883, brilliant twilight 
colors were seen around the world, as the dust gradually spread from 
its source until it encircled the globe. In mid-ocean, the air has been 
found to contain from 500 to 2,000 of these microscopic and sub- 
microscopic dust particles per cubic centimeter, and in dusty cities 
more than 100,000 per cubic centimeter. In the aggregate, large 
quantities of atmospheric dust are continually exchanged between 
the earth and the atmosphere. Current rain-making and weather- 
control theories now being tested are based on the importance of 
these hygroscopic dust particles to the weather processes. 

Properties of the Atmosphere 

By the "properties" of the atmosphere we mean the qualities or 
attributes of the air and its various constituents that contribute to 
the weather elements and atmospheric phenomena. Since the atmos- 
phere is essentially a mixture of gases, it behaves as other gases do- 
according to the natural laws of gas behavior, 

General characteristics. The principal characteristics of gases are 
their extreme mobility, compressibility, and capacity for expansion. 
The atmosphere is sometimes called the "ocean of air," and winds are 
compared to streams of water; but when these analogies are used, 
the greater freedom with which air moves in all directions, its greater 
fluidity and mobility, should be kept in mind. A gas has neither 
definite shape nor size. We cannot have a vessel half full of air; a 
small amount of air will fill a large vessel completely and uniformly. 
( Strictly speaking, exact uniformity is not attained, because the air 
at any point in the vessel is compressed by the weight of the air 
above it. This effect is negligible for most purposes in dealing with 
small volumes of air, but is of much more importance in considering 
the atmosphere as a whole. ) This property of indeterminate expan- 
sion is due to the fact that gases themselves exert a pressure which 
tends to change their volume to fit any container. This pressure is 


proportional to the density and the temperature of the gases and is 
exerted in all directions. According to the molecular theory, gases 
are made up of large numbers of minute molecules which are in a 
constant state of irregular motion. The effect of frequent collisions 
by these molecules may be observed in pressure and temperature 
characteristics. With this goes the property of great compressibility. 
The air is readily compressed; that is, its volume is decreased and its 
density increased when pressure is applied to it, and it readily ex- 
pands when the pressure is diminished. Under the same pressure, it 
expands with an increase of temperature and becomes denser with 
a decrease of temperature. In general, solids and liquids also change 
volume and density with change of temperature, but gases change 
to a much greater degree and with more uniformity. 

Laws of the gases. The characteristics just described are more 
definitely expressed in the following gas laws. These apply with close 
approximation to all the permanent gases of the air, but not so closely 
to water vapor. They refer to a fixed quantity (that is, mass) of gas. 

Boyle's Law. Robert Boyle ( 1627-1691), a British physicist and 
chemist of the seventeenth century, discovered that, for a given mass 
of a gas at constant temperature, the product of the pressure by the 
volume remains constant. (Fig. 3.) Stated otherwise, the volume 
of a given mass of gas varies inversely as the pressure on it if the tem- 
perature is unchanged. Algebraically, this may be written: 

PV = K, 

where P is the pressure, V is the volume, and K is a constant. A com- 
parison of pressure and volume under different sets of conditions 
may be expressed as: PiVi ~ PzVz PaVa, and so on. Since density 
means the ratio of the mass to the volume, density varies inversely 
as the volume, and therefore the density of a gas is directly propor- 
tional to its pressure, or P/D K' at constant temperature, where D 
is the density and K' is a constant. 

Law of Charles and Gay-Lussac. Two French physicists, Jacques 
Charles (1746-1823) and Joseph Gay-Lussac (1778-1850), discov- 
ered additional gas laws near the beginning of the nineteenth cen- 
tury. Charles' law states that when the volume remains constant, the 
pressure of a gas increases with the temperature at a rate of ^73 of 



PV = 2P X %F = 3P X VsF 
Fig. 3. Relation of Volume to Pressure on a Mass of Gas at Constant Temperature. 

the pressure at 0C. for each centigrade-degree increase in tempera- 
ture. Algebraically, this law may be expressed as follows: 

P* = Po [1 + ( M>73 ) T] at constant volume, 

where Ft is the pressure at any given time and Po is the pressure 
at 0C. 

Gay-Lussac showed that, at constant pressure, the volume of a gas 
increases with temperature at the rate of 1473 of its volume at 0C. 
for each centigrade-degree increase in temperature. This relation- 
ship may be expressed as : 

V* = Vo [1 + ( %73 ) T] at constant pressure, 

where Vt is the volume at any given time and V is the volume at 

One consequence of the first equation is that at a temperature of 
273C. a gas should cease to exert any pressure, that is, its mole- 
cules should cease to move. This temperature is called the absolute 
zero because, according to this law, no lower temperature can pos- 
sibly exist, since neither pressure nor movement can be less than 
zero. It has since been found that gases do not follow this law ex- 
actly at very low temperatures. Absolute temperatures are tempera- 
tures measured in centigrade degrees from the absolute zero. 

Combining the laws of Boyle, Charles, and Gay-Lussac, a single 


simple equation may be obtained expressing the relations among 
pressure, volume, and temperature for a given mass of any one gas: 

PV = RT, 

where T is the absolute temperature and R is constant for any one 
gas, but variable with different gases. 

Pressure. The air is held to the earth by the force of gravity. It 
therefore has weight, which is indicated by atmospheric pressure. 
At sea level, this pressure amounts to about 14.7 pounds per square 
inch, on the average, or about one ton per square foot. It decreases 
from this amount with increasing elevation and it also fluctuates 
slightly above and below 14.7 pounds per square inch at sea level 
with changes in the atmospheric conditions. 

The mass of air per unit volume is known as its density, dense air 
being a large quantity occupying a small volume. The weight of a 
cubic foot of air at sea level is about 1.2 ounces, or 0.08 pound. The 
density of the air is, therefore, 0.08 pound per cubic foot. Except for 
rare and temporary circumstances, the density and pressure of the 
air decrease with increasing height, rather rapidly at first and then 
more and more slowly. 

The Elements of Weather 

The first step in the development of a physical science is to ob- 
serve, measure, and record phenomena as they occur. From his early 
beginning, man has doubtless given attention to weather phenomena, 
but for thousands of years his observations were haphazard and were 
mere personal impressions, soon forgotten or distorted. Although 
some rainfall measurements were made at a very early date, it was 
only about three hundred years ago that man first began to measure 
the condition of the air and to record his observations for historical 
and comparative purposes. Such measurements necessarily awaited 
the invention of the thermometer and the barometer. The former is 
credited to Galileo (1564-1642) in about the year 1590; Torricelli 
(1608-1647) invented the mercurial barometer in 1643. 

Shortly after the invention of these instruments, some systematic 
observations of the temperature and pressure of the air were begun, 
but such observations did not become widespread, continuous, and 


comparable with one another until within the past 150 years. Today, 
when a meteorologist speaks of "making an observation" of the 
weather, he implies the careful use of instruments of precision for 
the purposes of determining and recording various physical facts 
about the condition of the atmosphere by methods so standardized 
that his observations are comparable with those of others through- 
out the world. His observations, however, are unlike those of the 
physicist in the laboratory; they are comparable rather to those of a 
physician in observing and measuring the condition of the patient. 
The constant and uncontrollable variations in the atmosphere are 
analogous to the changes in a living, moving organism. 

The meteorological elements. There are a number of physical 
properties and conditions of the atmosphere that may be measured 
quite accurately. Others must be observed or measured with less ex- 
actness because of the lack of mechanical measuring devices. All of 
these changeable properties must be measured accurately if we wish 
to determine what happens in the air and how it changes, or if we 
wish to describe the weather as it is at a given time and place. The 
most important of these are: (1) the temperature of the air, (2) the 
pressure that the air exerts, (3) the direction and speed of the air's 
motion, (4) the humidity of the air, (5) the amount of cloudiness, 
and (6) the amount of precipitation. These are the six fundamental 
weather elements. Other items are included in a complete weather 
observation, as will be noted later. It is evident that instruments 
giving results which can be set clown in figures are necessary for ob- 
taining an accurate knowledge of conditions and permitting a com- 
parison of weather at different times and places. In Chapters 2 and 3, 
the instruments and methods used in making the primary and essen- 
tial observations are described briefly. Instruments and observational 
details have been largely standardized in this country by the United 
States Weather Bureau. 

Weather and climate. Weather comprises the condition and char- 
acteristics of the atmosphere at a given time; climate implies the 
totality of weather conditions over a period of years. Climate is not 
merely the average weather; it includes also the extremes and vari- 
ability of the weather elements, for example, the greatest and least 
rainfall, the highest and lowest temperatures, and the maximum 
wind velocity for a given period. Since the weather is constantly 
changing, we need a long series of observations in order to have rea- 


sonably accurate information concerning the average and most fre- 
quent conditions and the probable variations. 

Many of the facts in reference to the weather and climate are ex- 
pressed in normal values. In meteorology, the word normal is used 
for the average, or mean, value of a weather element for a consider- 
able period of time. Ordinarily a mean value is not considered a nor- 
mal value unless there are at least ten years of record, and much 
longer records are required in most cases to establish an approxi- 
mately stable normal value. No such thing as an absolutely unchang- 
ing normal is known in climatic records. The normal, or average, 
value of a weather element is not necessarily its most probable value 
at any given time. 


Assuming the pressure of the air to be 14.7 pounds per square inch and 
its density to be 0.08 pound per cubic foot, solve the following problems: 

1. If air is removed from a vessel of 1 cubic foot capacity and of a total 
surface area of 800 square inches until the density is % that of the out- 
side air, what is the pressure on the vessel tending to crush it? 

2. What is the weight of the air remaining in the vessel? 

3. What volume of outside air is used to inflate an automobile tire to 
a pressure of 32 pounds per square inch in excess of the outside pressure, 
if the volume of the tube when inflated is 1,728 cubic inches? 

4. What is the density of the air in the tube? 

5. What is the weight of the air in the tube? 



Temperature Observations 

A weather element of primary concern is the temperature of the 
air. Temperature in many parts of the world is subject to wide ex- 
tremes and sudden changes; it is a weather element to which human 
life, and also plant and animal life, are sensitive; it is an important 
factor in determining the conditions of life and the productiveness 
of the soil in the different regions of the world; the varying tempera- 
ture of the air is responsible for many other weather changes. These 
are some of the reasons for the importance of temperature measure- 

Nature of heat and temperature. According to the molecular- 
expansion theory of the constitution of matter, all substances are 
made up of molecules in more or less rapid motion among them- 
selves. As the velocity of its intermolecular motion increases, the tem- 
perature of a body rises. Matter in motion possesses energy; it is ca- 
pable of exerting a force and of doing work; and the energy due to 
molecular motion is called heat . Heat is, therefore, a form of energy, 
and a measurable quantity, although not a substance. It may be 
transformed into other forms of energy. Although the human body 
is responsive to atmospheric temperatures, it is not an accurate in- 
strument for the measurement of the temperature of the air. For this 
purpose we need thermometers. 

Thermometers. Thermometers are instruments designed to re- 
spond accurately to changes of temperature. There are various types 



and forms of temperature-measuring instruments. They may be clas- 
sified on the basis of construction into four major groups: 1 

1. Liquid-in-glass thermometers which contain either mercury or 
some organic spirit such as ethyl alcohol or pentane. This is the most 
common type of thermometer for making surface weather observa- 

2. Deformation thermometers include the Bourdon thermometer 
( a curved, flattened, liquid-filled tube ) and the bimetallic thermom- 
eter which is actuated by the unequal expansion of two dissimilar 

3. Liquid-in-metal thermometers are variations of the Bourdon 
thermometer and are especially used by industry. The expansion of 
the liquid takes place in a separate, sealed container, and the change 
of pressure is transmitted through a small bore to the Bourdon tube, 
which may be located at considerable distance from the temperature 
being measured. 

4. Electrical thermometers are based on the change of the re- 
sistance to current as the temperature of the conductor is changed, 
or on the thermoelectric principle that when an electric circuit is 
made of two dissimilar metals and the junctions are not at the same 
temperature, a current will flow. Electric thermometers are not 
widely used for surface observations but are commonly used for 
upper-air measurements. 

The liquid-in-glass thermometer in common use for measuring the 
temperature of the air consists of a sealed glass tube with a small uni- 
form bore and an expanded bulb at one end. The bulb arid a portion 
of the tube are filled with a liquid, usually mercury. The height of 
the liquid in the tube changes as the temperature of the mercury 
changes, because mercury, like other substances, expands as its tem- 
perature increases and contracts as its temperature decreases. The 
change in volume of the mercury is proportional to the change in 
temperature. The glass of the thermometer responds to temperature 
changes also, but the coefficient of expansion of mercury is about 
seven times that of glass. 

The height of the liquid in the tube, when the thermometer is at 
the temperature of melting ice, is accurately determined and marked 
upon the glass, and similarly a point is marked indicating the tem- 

1 W. E. K. Middleton, Meteorological Instruments (Toronto: University of Toronto 
Press, 1947), pp. 55-90. 


perature of boiling water under standard conditions of pressure. 
When these two "fixed points" have been determined, the distance 
between them on the tube is divided into a number of equal divi- 
sions, called "degrees," and the divisions may be extended beyond 
these two points in each direction. The length of the bore and the 
amount of mercury in the tube are determined by the range of 
temperature conditions expected to be measured by a given ther- 

On the Fahrenheit thermometer, invented in 1710 by Daniel Fahr- 
enheit (1686-1736), a German physicist, the temperature of melting 
ice is called 32 and that of boiling water, 212. Although invented 
by a German scientist, the Fahrenheit thermometer is now in com- 
mon use only in English-speaking countries. On the centigrade scale, 
the freezing and boiling points of water are called and 100, re- 
spectively. The centigrade thermometer is also sometimes called the 
Celsius thermometer, after the Swedish astronomer, Anders Celsius 
(1701-1744), who invented it in 1742. Temperatures below the zero 
of either scale are written with a minus sign. A change in tempera- 
ture from 32 to 212, being a change of 180 on the Fahrenheit 
scale, corresponds to a change of 100 on the centigrade scale, 
making each Fahrenheit degree equal % of a centigrade degree 
(Fig. 4). For some scientific purposes, it is preferable to use a scale 
which has its zero at 273C. and ascends in centigrade units. This 
is called the absolute scale and is indicated by the letter "A" follow- 
ing the number. The following formulas may be used to convert from 
one scale to another: 

C=% (F-32) ^A-273, 
F = % C + 32, 
A = C + 273. 

An accurate thermometer meets the following requirements: the 
bore is uniform, the fixed points are accurately determined, and the 
graduations are correctly spaced and etched on the stem. It contains 
a suitable fluid, one that does not freeze at the temperatures to be 
measured and does not readily vaporize or decompose. To meet the 
requirement of a nonfreezing liquid, alcohol is used instead of mer- 
cury under very cold conditions, for mercury freezes at 38.7F 
(~39.3C). The size and shape of the bulb and the size of the bore 


Fahrenheit Centigrade Absolute 


f>1 *> 

4 f\f\ 

373 -*- Point of 

368 Water 






























































278 Melting 
273 - Point of 

268 Ice 





















> i 




238 Freezing 
233 point of 
228 Mercury 


Fig. 4. Thermometer Scales Compared. 

determine the instrument's sensitiveness and quickness of response. 
Maximum thermometers. Special thermometers are used for ob- 
taining the highest and lowest temperatures occurring during any 
interval. The maximum thermometer has a constriction of the bore 
just above the bulb, through which the mercury is forced out of the 
bulb as the temperature rises but does not flow back as the tempera- 
ture falls (Fig. 5). The top of the column therefore remains at the 
highest point reached since the last setting of the thermometer. The 


maximum thermometer is set by whirling it around a mounting near 
its upper end. The centrifugal force thus generated forces the mer- 
cury back into the bulb. After setting, the thermometer indicates the 
correct temperature at the time, called the current temperature. The 
maximum thermometer should be mounted nearly horizontal with 
the bulb slightly higher than the other end to lessen the tendency for 
the mercury to retreat into the bulb. 

Fig. 5. Maximum and Minimum ThcrmomeU'is with Towiiseml Support. Courtesy, 
Friez Instrument Division, Bendix Aviation Corp., Baltimore, Md. 

Minimum thermometers. The liquid used in a minimum ther- 
mometer is alcohol. A small, dumbbell-shaped, glass index is placed 
within the bore of the instrument. The thermometer is mounted hori- 
zontally, with the index within the liquid and in contact with its 
surface at the end of the column. As the temperature falls and the 
column shortens, the index is carried toward the bulb by the surface 
tension of the liquid. When the temperature rises, the liquid flows 
past the index and leaves it at the lowest temperature reached. After 
the minimum temperature is read, the index is returned to the top 
of the liquid, which is the current temperature, by simply turning 
the thermometer bulb-end up. By making readings of a maximum 
and a minimum thermometer once a day, the highest and lowest 
temperatures reached during the 24 hours are obtained. A reason- 
ably regular reading-hour should be established that is not likely to 
coincide with either of the extreme temperatures. 

Thermographs. Various types of recording thermometers, or ther- 
mographs, are vised to obtain a continuous record of the tempera- 
ture. One common type uses a Bourdon tube, which has a flattened, 
curved metal tube filled with liquid, sealed, and fastened rigidly at 



one end. With change of temperature, there is unequal expansion 
or contraction of the liquid and the metal, producing a change of 
curvature in the tube, thus moving the free end. This movement is 
communicated to a pen, which is caused to move up or down on a 
drum that is being slowly rotated by a clock within it. In this man- 
ner, a continuous record of the temperature is traced on a ruled 
sheet surrounding the rotating drum (Fig. 6). Such a record is less 
accurate than one obtained by readings of a mercury thermometer, 
but if the thermograph trace is checked and corrected occasionally 
by comparing it with an accurate and similarly exposed thermometer, 
the results are sufficiently accurate for general meteorological pur- 

Fig. 6. Thermograph. A clock mechanism rotates the cylinder beneath the re- 
cording pen. Temperature changes cause changes in the curvature of the Bourdon 
tube, A, resulting in vertical movements of the arm, B. Courtesy, Friez Instrument 
Division, Bendix Aviation Corp. 

Obtaining the temperature of the air. To determine the air's tem- 
perature, more is required than an accurate thermometer. It is 
equally important to make sure that the thermometer assumes the 
temperature of the air. A thermometer indicates its own temperature, 
but sometimes that is not the same as the temperature of the air sur- 
rounding it. If the thermometer is exposed to direct sunshine or to 



reflected heat from ground or buildings, it becomes hotter than the 
air around it. If it is close to a good radiating surface at night, it 
becomes colder than the air. If it is exposed where the air does not 
move freely, it may indicate the temperature of the air immedi- 
ately around it but not of the general mass of air. These are some 
of the reasons why even good thermometers disagree. Most privately 
owned thermometers are not properly exposed, yet this is just as im- 
portant as having a good thermometer. There often are actual dif- 
ferences in air temperature within short distances, and thermom- 
eters in the same city should not necessarily agree; but in many cases 
the disagreement is due to a failure to secure the temperature of the 
free mass of air. 

To obtain the correct temperature reading, thermometers are ex- 
posed to the freely moving air in such a way that they are screened 
or sheltered from other influences. The instrument shelter in stand- 
ard use in the United States is a white box with a base about 2 by 
2% feet, and about 33 inches high (Fig. 7). It has a sloping double 


Fig. 7. Instrument 

iter with Door Open and Instruments in Place. Courtesy, 
U. S. Weather Bureau. 


roof with open air space between. All four sides are louvered to 
permit free movement of air through it while protecting the instru- 
ments from sunshine, rain, and snow. The bottom is nearly closed 
but permits some movement of air through it. It is preferably 
mounted over sod, about four feet above the ground, to get above 
the influence of the surface temperature and into a layer of air that 
is moving freely. Shelters embodying the same principles are used in 
meteorological services throughout the world. 

Uses of temperature observations. Standardized observations of 
temperature have been made in some places in Europe for more 
than 100 years, and in the United States for more than 60 years. 
Many stations in each state have records longer than 35 years. For 
stations having sufficient length of record, normal annual, monthly, 
and daily values may be obtained; also, the mean maximum and 
minimum temperatures and the actual extremes of highest and low- 
est temperatures are determined. Monthly and annual temperature 
normals based on 10 years of record are frequently used, but 30 to 
40 years give more trustworthy normals. Hourly values may be ob- 
served and recorded by an observer or, more frequently, read from 
the thermograph sheets in stations equipped with this instrument 
(Fig. 8). The mean temperature for a given day may be obtained 
by taking the mean of the 24 hourly readings, but the sum of the 
maximum and minimum divided by 2 is generally used instead. The 
mean temperature of a given month is the mean of the average 
maximum and average minimum temperatures for that month. The 
annual mean temperature is the average of the 12 monthly means. 

From the hourly values, or the thermograph record, the daily 
march of temperature may be learned, by which we mean the regu- 
lar progress of temperature between low and high points during the 
day. On the average, the highest temperature for the day occurs, 
not at noon, but in mid-afternoon, between 2 and 5 P.M. Most heat is 
received from the sun at noon, but during a portion of the afternoon, 
the earth and the air near it continue to receive more heat than they 
lose, and hence the temperature continues to rise until a balance 
between incoming and outgoing heat is reached. This delay in the 
occurrence of the maximum until a few hours after noon is known 
as the retardation, or lag, of the maximum. From the time of the 
maximum, the temperature usually falls rather rapidly until about 8 




to 10 P.M., and then more slowly until additional heat is again re- 
ceived from the sun. The time of the minimum is therefore just be- 
fore sunrise. These are average conditions; on any one day there 
may be irregular fluctuations which upset this regular march of tem- 
perature. The difference between the highest and lowest tempera- 
tures for any day is called the daily range of temperature. Different 
daily ranges indicate important climatic differences. For example, 
the average daily range at Key West, Florida, is about 10F., and at 
Wennemucca, Nevada, about 30F., indicating that Key West, Flor- 
ida, has little change in temperature from day to night, and Winne- 
mucca, a large change. 

The annual march of temperature in most parts of the Northern 
Hemisphere makes January the coldest month and July the warm- 
est. The reverse is true in the Southern Hemisphere. In the interior 
of the United States, daily normals of temperature reach a maximum 
about July 15-25 and a minimum about January 15-25, but the most 
heat is received on June 21-22 and the least on December 21-22. 
There is thus a retardation of both maxima and minima of about 
one month (Fig. 9). Where temperatures are influenced by large. 







* "* 




s^ 1 ^ 









N * 

^ ^ 







^ ^ 










L V 




\ % 











_.. ^ 












^ ^_ 


Fi. 9. Typical Curves Showing Annual March and Annual Range of Temperature. 

bodies of water, the retardation is often greater than one month. In 
middle latitudes, individual years are marked by great irregularity 
in the march of temperature; that is, by irregularly alternating spells 
of warm and cool weather of unequal length, so that in a given year 
June or August may be warmer than July, and December or Febru- 



ary colder than January. The annual range of temperature means the 
difference between the mean temperature of the warmest and the 
coldest month. 

Pressure Observations 

The pressure of the air at a given place is a force exerted in all 
directions in consequence of the weight of all the air above it. As 
a result of the air's constant and complex movements and the changes 
in its temperature and its water-vapor content, the weight of air 
above a fixed point is continually changing. The pressure, therefore, 
like the temperature, is never constant for long; but, unlike tem- 
perature changes, variations in pressure are not ordinarily percep- 
tible to human senses. They are, nevertheless, an important feature 
of the weather by reason of their relations to other weather changes. 
Mercurial barometers. The instrument used to measure the at- 
mospheric pressure accurately is the mercurial barometer. When a 
glass tube about three feet long is filled with mercury and then in- 
verted and the open end immersed in a cup of mercury, the mer- 
cury will flow out of the tube into the cup until the weight of the 
column in the tube ( above the surface of the mercury in the cup ) is 
balanced by the pressure of the air upon an equal cross section of 
the liquid surface (Fig. 10). The length of the column of mercury 

thus becomes a measure of the 
f^l VACUUM pressure of the air. This is the in- 

U strument invented by Torricelli in 

H 1643. The instruments in use today 

are only mechanical refinements of 

the original barometer. It becomes 
H obvious that fixed gradations on the 

tube for measurement of the col- 
H umn height would not be satisfac- 

MERCURY t ol y un less a method were devised 
n / n to keep the height of the mercury 
^l in the cup, or cistern, at a constant 
[BBHHBBBI level. In the Fortin type of mer- 
cury barometer, the cistern has a 
flexible bottom to which is at- 
tached an adjusting screw by which the level of the mercury in 
the cistern is set at a fixed point before each reading (Fig. 11). 

Fig. 10. A Simple Mercurial 



Fig. 11. Mercurial Barometers Mounted in Case. From left to right, Fortin type 
with screw to adjust level of mercury in cistern, fixed-cistern type, and a barometer 
combining principles of the other two. Courtesy, U. S. Weather Bureau. 

A thermometer is attached to the frame of the barometer for rea- 
sons to be explained later. 


Units of pressure measurement. In this country, the barometer 
scale is usually graduated in inches. In countries where the metric 
system is in use, the scale is marked in millimeters. When we say 
that the barometer reads 29.92 inches or 760 millimeters, we mean 
that the pressure of the air supports a column of mercury of that 
length. This value, 29.92 inches or 760 millimeters, is taken as the 
normal value of the pressure at sea level at latitude 45, and is called 
the normal atmosphere, or simply one atmosphere. There has now 
come into general use another unit of atmospheric pressure, called 
the bar, which is not a measure of length but a direct statement of 
force per unit area, that is, of pressure. Barometer scales are fre- 
quently marked in millibars (mb) thousandths of a bar. The bar is 
equal to 1 megadyne (1,000,000 dynes) per square centimeter. 
Under standard conditions of temperature and gravity, a pressure 
of 29.53 inches = 1 bar = 1,000 millibars. The millibar, as a unit of 
measure of atmospheric pressure, is in widespread use among the 
weather services today, and pressures are seldom converted to inches 
of mercury except for public uses. A comparison of the three scales 
for measuring atmospheric pressure is made in Fig. 12. Note that 
1013.2 mb = 29.92 in. = 760 mm. 

Barometer corrections. The length of a mercury column which 
a given pressure of air will support depends upon the density of the 
mercury, and this changes with the temperature. Therefore, to make 
an accurate pressure reading, we must take into account the tem- 
perature of the mercury; and to compare readings at different times 
or places, we must make a temperature correction. For this purpose 
a thermometer is attached to the barometer, and all readings are cor- 
rected to what they would be at a standard temperature. The stand- 
ard in use for the temperature of the mercury is 32 F. The force of 
gravity varies over the earth's surface, decreasing from the poles to 
the equator. This is so because the earth is not a perfect sphere, but 
a spheroid whose equatorial diameter is about 27 miles (43 km) 
greater than the polar diameter. Hence, the same actual pressure 
would raise the mercury higher at the equator than at the poles, be- 
cause at the equator it weighs less; and it is necessary to make a cor- 
rection for gravity depending on the latitude of the barometer. 

Each barometer, before being put into use, should be compared 



with a standard precision instrument. Each is usually found to have 
certain divergences due to scale inaccuracies and capillarity, and 

these are grouped together under the 
name of instrumental errors. In a well- 
made instrument, the instrument errors 
are less than 0.01 inch. When these cor- 
rections are applied to the observed 
barometer reading, the result is the sta- 
tion pressure, that is, the pressure at a 
definite place and time. A correction, or 
adjustment, for the altitude of the ba- 
rometer above mean sea level is usually 
made by the observing station so that 
the pressures of stations at different ele- 
vations may be comparable. 

Aneroid barometers. Another instru- 
ment in general use for the measurement 
of pressure is the aneroid barometer. It 
consists essentially of a flexible metal 
box, or chamber, which is hermetically 
sealed after being nearly exhausted of 
air, and is kept from collapsing by a 
spring within it. The flexible chamber 
then responds sensitively to pressure 
changes, and the resulting movements 
are communicated to an index hand mov- 
ing over a dial ( Fig. 13a ) . Aneroid ba- 
rometers are compensated for tempera- 
ture and require no gravity correction; 
the station pressure is read directly from 
the dial. Instrumental errors, however, 
are considerable and variable, and these 
instruments are less reliable than are 
mercury barometers and should be 

checked frequently with them. Aneroids are light and easily car- 
ried without injury, if not subjected to severe jarring, and are there- 
fore useful for travelers and explorers and on vessels at sea (Fig. 




* 1 AA 



*A AA 


I ni ft 



9Q *X 


7t A 


1 A AA 

O<\ AA 

7 OO 






71 1 

9oZ "~ 


97 AA 

/ A 

O 1 A 

'" Z7.OO 

" -- . 

1 IN * 25. 
1 MM* 1.3 


p*-~ ""*-* 

4 MM 33 


"s- *_ ^ 

36 MB 

Fig. 12. Barometer Scales 

Fig. I3i\. Schematic Drawing of an Aneroid Barometer. 

Fig. 13b. Navy Type Aneroid Barometer Calibrated in Inches and Millibars. 
Courtesy, Friez Instrument Division, Bendix Aviation Corp. 



A barograph is an aneroid barometer that makes a continuous rec- 
ord of the pressure. It consists of several metallic chambers, one on 
top of the other. The combined motion of these is communicated to 
a lever, terminating in a pen. The pen writes a record of the pressure 
upon a ruled sheet of paper wound around a drum while the drum 
is being rotated slowly by a clock within it (Fig. 14). The continu- 
ous records of pressure thus obtained are valuable in showing the 
march of pressure, the extremes, and how the pressure is varying at 
any time. The barometric tendency, meaning the change in pressure 
during a given period of time (usually, the three hours preceding 
an observation) is of importance in forecasting the weather. 

Fig, 14. Barograph. Ihe flexible, metallic chambers, A, are partially exhausted 
of air and a system of internal springs are balanced against the outside* air pressure. 
Changing air pressure results in compression or expansion which causes a vertical 
movement of the pen. Courtesy, Friez Instrument Division, Bendix Aviation Corp. 

Variation of pressure with height. As we rise above sea level, we 
get above some of the weight of the air, and the pressure falls rather 
rapidly at first in the dense lower air, and then more slowly as the 
air becomes thinner. As a first approximation, we may say that the 
pressure decreases l/30th of its value at any given moderate alti- 
tude with an increase of 900 feet in height (275 m). Starting with 
a pressure of 30 inches at sea level, at 900 feet above sea level it will 
have fallen to 29 inches; during the next 900-foot rise to 1,800 feet 
elevation, it will have fallen l/30th of 29 inches to 28.03 inches; con- 
tinuing at this geometric ratio for each successive change of 900 


feet. But the density and weight of the air depend upon its tempera- 
ture and, to a lesser extent, upon the proportion of water vapor in 
it and the force of gravity. Hence, no accurate correction for alti- 
tude can he made without a consideration of these factors, espe- 
cially the temperature. 

Reduction to sea level. In studying the distribution of pressure 
over the earth, it is necessary, then, to take account of the differing 
altitudes of the places at which the pressure was measured. In doing 
this, all readings are customarily "reduced to sea level." For places 
ahove sea level, this means adding to the station pressure an amount 
assumed'to represent the weight of the air in a vertical column ex- 
tending from the point of observation to sea level. Since no such 
column exists beneath a land station, assumptions as to its tempera- 
ture, density, and moisture content are fictitious, and the results are 
only approximations. When the altitudes are considerable, as in the 
Rocky Mountain region, the reductions thus made are subject to con- 
siderable error. 

The various corrections to be applied in reducing pressures to sea 
level, or conversely, in determining heights by the barometer, are 
published in detail in the Smithsonian Meteorological Tables. By ap- 
plying these corrections, the difference in altitude between two near- 
by places may be determined with considerable accuracy if simul- 
taneous observations of pressure, temperature, and humidity are ob- 
tained at the two stations. 

Altimeter. The relation of pressure to height above sea level has 
long been used by travelers, explorers, and surveyors in making esti- 
mates of altitudes and differences in altitude, and now has a wide 
application in connection with aviation. The pressure altimeter, car- 
ried by all airplanes, is an aneroid barometer graduated to read 
directly in heights instead of pressures. The rate of decrease of pres- 
sure with height varies with the temperature of the air column and 
to a lesser degree with the pressure distribution and the humidity of 
the air. The elevations indicated by such an altimeter are correct 
only under the assumed standard conditions. If the air is colder than 
the standard atmosphere, the instrument indicates too high an alti- 
tude; if the air is warmer than the standard, the reading is too low. 
Another source of error lies in the fact that the pressure differs widely 
from time to time at the same place, and also differs from place to 
place at the same height above sea level. 


Before taking off on a flight, the pilot adjusts his altimeter to an 
altimeter setting, which is the pressure at the airport reduced to sea 
level by assuming standard atmospheric conditions, not by using 
current temperatures. His instrument will then read the correct alti- 
tude of the airplane at that moment. As the time since the altimeter 
was set increases, and as the plane gets farther from the place where 
it was set, the probability of serious error in the indicated altitude 
increases. The error may amount to several hundred feet. Obviously, 
it is wise to check the altimeter setting frequently with points along 
the flight route, and especially to adjust the instrument to the cor- 
rect altimeter setting at the destination before landing. An instru- 
ment developed during World War II, using the radar principle, is 
called the radio altimeter. It is installed on all of the larger aircraft 
today and enables the pilot to make a direct determination of his 
true elevation above the ground and is therefore an absolute alti- 

Results of pressure observations. Pressure observations began in 
Italy about the middle of the seventeenth century and have been 
carried on more or less continuously in various parts of the world 
from that time to the present. Especially during the past hundred 
years, observations have been numerous and widely distributed. 
Yearly, monthly, daily, and hourly normals have thus been estab- 
lished, more or less definitely, throughout the world. The normal 
annual pressure is found to differ in different parts of the world, and 
the monthly normals at any one place change with the seasons. Out- 
side the tropics, there are also comparatively large irregular varia- 
tions from day to day, independent of seasonal changes but more 
marked in winter than in summer. 

Diurnal variations. Finally, there are regular daily variations of 
small amounts, resulting in two maxima and two minima each day. 
The maxima occur about 10 A.M. and 10 P.M. local time, and the 
minima about 4 A.M. and 4 P.M., varying somewhat with the season 
of the year. These diurnal variations are greatest in equatorial re- 
gions, where they amount to about 0.1 inch (3 mb), and grow stead- 
ily less toward the poles. In higher latitudes, they are practically 
masked by the larger irregular variations and may not be apparent 
except in the averages of a long period. A complete physical ex- 
planation of these daily changes is difficult, but they seem to be long 
atmospheric waves similar to the ocean tides, which move around 


the earth about two hours in advance of the sun and are in a com- 
plex way associated with the gravitational attraction of the sun and 
the daily changes in temperature. 

Wind Observations 

Wind is air in horizontal motion. Vertical movements in the air 
are commonly called currents. Winds are of fundamental importance 
in making our weather what it is. In the first place, the motion it- 
self is a weather factor of importance a quiet winter's day may be 
pleasant and a windy day may be disagreeable. In the second place, 
the physical condition of the air is largely a function of its source 
and horizontal movement. Winds become moist when moving over 
large water areas, and they carry this moisture to the land. Air be- 
comes cold over frozen or snow-covered regions and moves, as wind, 
to warmer regions. Similarly, warm air is transported to normally 
cold regions. To describe the movement itself without reference to 
the condition of the moving air, two facts about the wind must be 
observed, namely, its direction and its speed. 

Wind direction. Wind vanes have been in use since ancient times 
as indicators of the direction of the wind. A wind is named for the 
direction from which it comes, that is, the direction toward which 
the arrow of the wind vane points (Fig. 15). Winds are said to veer 
when they change in a clockwise direction, such as east to south 
or west to north, and are said to back when they change in the op- 
posite order. 

In order to secure a sensitiveness of response and at the same time 
a comparative steadiness in gusty winds, a wind vane about 30 inches 
in length may be used, mounted on roller bearings and having a tail 
of two pieces, each 8 inches wide and 12 inches long, making an 
angle of 22 with each other. An automatic record of the wind di- 
rection at intervals of one minute may be obtained by attaching a 
cam collar to this vane and connecting it by electrical circuits to a 
recording device actuated by clockwork. A recording instrument in 
common use is the meteorograph or triple register. It records not 
only wind direction, but also wind speed, sunshine, and rainfall, as 
will be noted later. Hence the name "triple register," because it re- 
cords three weather elements. 

Other devices are in use, called anemoscopes, which give a con- 



Fig. 15. Wind Instruments. These instruments are mounted on the tower support 
on top of the observatory of the central office of the U. S. Weather Bureau, Wash- 
ington, D.C. From left to right, Dines (pressure tube) anemometer, thunderstorm 
indicator (generating voltmeters), 4-foot wind vane, and 3-ciip anemometer. The 
anemometer is whirling quite rapidly so that the cups are not discernible. Courtesy, 
U. S. Weather Bureau and L. E. Johnson. 



tinuous record of wind direction. There are also wind-direction in- 
dicators showing exact directions at any instant by a pointer on a 
dial. Both surface winds and winds aloft are observed according to 
36 points of the compass, or to the nearest ten degrees (Fig. 16). 

Kitf. 10. Wind Directions. 

Wind direction is plotted on a weather map by means of a wind 
shaft drawn toward the station from the direction of the wind. A 
surface wind is plotted as W, WNW, NW, NNW, and so forth, and 
velocity is shown by feathers on the shaft (Fig. 17). Winds aloft are 
represented in the same manner with accuracy to 36 points of the 

Wind speed. Moving air exerts a force or pressure against ob- 
jects in its path, and that force is proportional to the square of its 
velocity. 2 This may be expressed by the equation: P = K\ r \ where P 
is the pressure exerted by the wind, V is its velocity, and the value 
of K depends upon the units used. If pressure is expressed in pounds 

2 In strict usage, velocity is a vector quantity and equals speed in a particular di- 
rection. In common usage and as ordinarily tabulated, wind velocity refers to speed 
of motion without reference to direction. 



per square foot and velocity in knots* then P 0.0053V 2 , approxi- 
mately, for a flat surface normal to the wind, where pressure includes 
also the suction on the rear of the surface. (When velocity is ex- 
pressed in miles per hour, the formula becomes P = 0.004V 2 . ) By rea- 
son of this force exerted by the wind, its veloctiy can be estimated 
without instruments by its effect on surrounding objects. For this 
purpose a scale has been developed, known as the Beaufort scale. 
Originally developed by Admiral Beaufort of the British Navy in 
1805, this scale was at first expressed in terms of the effect of the 
wind on the sails of a ship but has since been adapted for both land 
and sea use ( Fig. 17 ) . 



Miles Per 

Map Symbol 
(Northeast Wind) 


Less than 

Less than 








Light Air 




Light Bree/e 





Gentle Bree/e 




' ' /= 

Moderate Bree/e 





Fresh Bree/e 





Strong Bree/e 





Moderate Gale 





Fresh Gale 




/^ f 

Strong Gale 





Whole Gale 














Fig. 17. lit 

lufort Wind Sc 

ale. A simplifie 

d scale rapidly gaini 

ig popularity is one 

where each half-feather represents five knots of wind and fifty knots are represented 
by a triangular pennant. 

For a mechanical measurement of wind velocity, several types of 
anemometers have been developed. A simple deflection anemometer 
consists of a board, hinged at the top and swinging in the wind, hav- 
ing an attached arc to indicate the angular amount of its deflection 

3 The knot, one nautical mile (6,080.20 feet) per hour, has become the standard 
unit of wind velocity in the United States, The term originated at sea from divisions 
in the log line arranged to measure a ship's speed through the water. 



from the vertical. From this deflection the velocity may be calcu- 
lated. A pressure tube anemometer consists of a U-shaped tube con- 
taining a liquid and having one of the open ends directed toward 
the wind. The difference in level of the liquid in the parts of the 
tube is a measure of the pressure of the wind and hence of its 

Robinson cup anemometer. For meteorological purposes, the 
Robinson cup anemometer has long been in general use in this coun- 
try. In this type of instrument, a set of hemispherical cups is mounted 
on a vertical axis attached to a spindle which actuates a dial (Fig. 
18). As the cups revolve in the wind, distances are indicated on the 
dial in knots or miles per hour. In making such an instrument it is 
necessary to use a fixed ratio between the speed of the cups and the 
distance indicated on the dial. In fact, however, the ratio of the 
speed of the cups to the true velocity of the wind varies, making the 
readings of the anemometer too small at low velocities and too great 
at high velocities. This has been determined by careful calibration 
of such anemometers in a wind tunnel, where the movement of the 
air is known accurately by other means. The corrections so deter- 
mined are applied to the indicated velocities before publication. 

By fitting the dial with posts that 
press against a spring and thus 
close an electric circuit for each 
mile, an automatic record of the 
wind movement is obtained on the 
meteorograph on the same sheet 
with the record of the direction. 
Both the anemometer and the wind 
vane may also be connected elec- 
trically with an indicator located in 
an office at some distance in such 
a manner as to enable the observer 
to determine the direction and 
speed of the wind at any time with- 
out visiting the instrument. 

Aerovane. A recent develop- 
ment is the Bendix-Friez aerovane 
wind transmitter, a combined anemometer and wind vane. It uses a 
three-bladed propeller for measuring the speed of the wind, and a 

Fig. IB. Holunxon 3-cnp Anemom- 
eter. Tottili/m.u; dials indicate the total 
miles of wind, and electrical contacts 
operate a speed indicator and recorder. 
Courtesy, Friez Instrument Division, 
Bendix Aviation Corp. 



Fig. 19. Aerovane. Instantaneous direction and speed of the wind may be read 
remotely from the dials connected to this instrument. Courtesy, Friez Instrument 
Division, Bendix Aviation Corp. 



streamlined vane for direction (Fig. 19). The propeller rotates at a 
rate proportional to the wind speed. The vane performs two func- 
tions: it indicates the wind direction, and it keeps the propeller axis 
pointed into the wind (Fig. 18). Both are connected with indicat- 
ing or recording instruments. 

Custiness of winds. The record made by a cup anemometer 
gives the time between successive miles of wind, but a pressure tube 
anemometer may be arranged to give a continuous graph of the 
fluctuations of the wind. Other types of instruments to indicate the 
gustiness and instantaneous speed of the wind have been devised. 
One of the simpler and more practical of these is the Gurley electric 
anemometer, a three-cup anemometer in connection with a high- 
frequency oscillator. A recording mechanism may be attached for 
making a continuous, permanent graphic record. The records ob- 
tained by such instruments show that the flow of air near the surface 
of the earth is never steady. It is not a streamline flow, but a move- 
ment in successive gusts and lulls of a few seconds' duration (Fig. 
20). This turbulence is greater the higher the wind velocity; it is 
greater over land than over ocean surfaces, and greater over forests 
and cities than over bare, level ground. Evidently, the turbulent 
motion is caused, in part at least, by surface irregularities and 

Fig. 20. AnriiioM'opi- Krcoul, Slum ing Uu.sUiuv> of the \Vind. Courtesy U. S. 

Weather Bureau. 

Friction at the earth's surface induces gustiness by checking the 
flow of the lowest layer, letting the layers above it break over it like 


the waves along a sloping seacoast. Surface obstacles turn the air 
out of its course and into numerous cross currents. Eddies around 
buildings and through city streets are familiar examples of turbulent 
motion, but all the trees and shrubs and all the little irregularities 
of the land cause similar eddies in relation to their size, changing 
both the speed and the direction of the wind in their vicinity. The 
effect of such obstructions extends to five or six times their altitudes. 
These effects are, therefore, local and confined to the air near the 
earth's surface, unless other forces aid in causing unsteady motion. 

Effect of altitude. The average velocity of the wind increases 
with height above the ground. There is a marked increase in the first 
100 feet (30 m). In general, the velocity at the height of 33 feet is 
about twice that at l 1 ^ feet, and the velocity at 100 feet is 1.2 times 
that at 33 feet. This reduced velocity near the surface is evidence of 
the "frictional drag" of the earth. Notwithstanding the increased 
velocity, there is less turbulence as we rise into the free air, but some 
effects of surface eddies are felt to heights of 6,000 to 9,000 feet 
(1800-2700 m). Turbulence also sometimes originates in the upper 
air through the contact and resulting friction of winds of different 
directions or velocities and different densities. 

Exposure of wind vane and anemometer. Considering the effects 
of surface turbulence and surface drag, it is evident that, in order to 
get records truly representative of the general conditions in a region, 
the place of exposure of the wind instruments must be carefully se- 
lected. They should be placed where they are as free as possible of 
interference from local irregularities, that is, as far as possible from 
adjacent high objects and as much as possible above them. Often 
they are placed on the roofs of buildings at varying distances from 
othej buildings of comparable height, and are raised on steel sup- 
ports from 15 to 25 feet above the roof, thus making the elevation 
above ground 50 feet (15 m) or more, and sometimes 200 or 300 
feet. Exposure of anemometers on the roofs of buildings and at vary- 
ing elevations in different localities is unsatisfactory for comparative 
purposes. It would be better if all could be exposed on towers in 
the open and at standard heights above the ground. Exposures at 
airports are generally more satisfactory than at city offices. Valleys, 
even shallow ones, affect the direction at the surface markedly and 
the speed to a less extent, and the wind records obtained in moun- 
tainous regions are seldom representative of large areas. 


Results of wind observations. The records obtained by continued 
observations of the wind afford valuable information which may be 
summarized in various ways. Official records in the United States 
give the prevailing direction and the average velocity for each day, 
month, and year, and the monthly and annual normals; also, the 
maximum velocities by months, and the number of days when veloci- 
ties of 32 miles per hour ( 16 mps ) or more occurred. Also calculated 
are the percentages of the time that the wind blew from each of the 
principal directions and the percentages of the total movement from 
each direction. Wind data may be graphically presented by means 
of a wind rose, in which the relative lengths of the radiating lines in- 
dicate the relative frequency of the winds from the different direc- 
tions (Fig. 21). 

Annual variations. The accu- 
mulated observations show that 
there is an annual change in both 
the direction and the speed of the 
wind in most parts of the world. 
The velocity is greater, on the aver- 
age, in winter and spring than in 
summer and autumn. The reason 
for this is the greater contrast in 
temperature between high and low 
latitudes in winter and spring sea- 
sons, as will be noted later. Usually 
March is the month of highest 
average velocity and August the 
month of the lowest; but in a large 
part of the Mississippi and Mis- 
souri valleys, April is windier than 

March. In Rocky Mountain and Pacific Coast regions and in the 
vicinity of the Great Lakes, there is considerable local variation in 
the months of highest and lowest average velocity. The prevailing 
direction of the wind also frequently changes with the seasons, 
owing to changing temperature contrasts between land and ocean 

Diurnal variations. The velocity of the wind is generally greater 
by day than by night over land surfaces, especially in summer and 


Fig. 21. Wind Rose for New York 
City. Average annual percentage of 
winds from eight directions. 


on clear days. The highest average occurs from 1 to 3 P.M., and the 
lowest about sunrise. These diurnal variations are caused by the 
heating and rising of the surface air by day and the descent of cooler 
air from aloft. This explanation is confirmed by the fact that at sea, 
where the surface does not become diurnally heated, there is little 
difference between day and night velocities. The conditions under 
which vertical interchanges of air take place are discussed in Chap- 
ter 5. In most coastal regions, there is a daily change in wind direc- 
tion, which will be considered in more detail in Chapter 7. 

Irregular variations. Although the annual and daily variations of 
the wind movement recur with more or less regularity, they are sub- 
ject to continual interruption by irregular changes due to special 
caiuses. Sometimes these erratic variations occur as squalls sudden 
marked increases in velocity, like gusts but lasting much longer. At 
other times, the wind may shift radically in both direction and speed 
and may continue from the new direction for several hours or even 
days. An explanation of these irregular wind variations must be de- 
layed for later discussion. 


1. Express the following Fahrenheit temperatures in the centigrade 
scale: 86; 44; 23; -13. 

2. Change the following centigrade temperatures to Fahrenheit: 35; 
23; 10; -10; -20. 

3. The following barometer readings are given in inches: 28.75; 29.54; 
30.15; 30.36. Express them in millimeters and in millibars. 

4. What is the approximate barometric pressure in inches and in milli- 
bars at the following elevations: 1,800 feet; 2,700 feet; 1 mile? 

5. What is the pressure in pounds per square inch at the elevations in 
problem 4? (A cubic inch of mercury weighs 0.49 pound.) 

6. Calculate in round numbers the total weight (pressure at the sur- 
face ) of the earth's atmosphere. 

7. From data given in Table V, Appendix 3, draw graphs of the an- 
nual march of temperature at Honolulu, London, Moscow, Freetown, 
Melbourne, and note variations in annual range and in retardation of 
maximum and minimum. Compare with Fig. 9. 

8. What pressure does the wind exert against a wall, 60 by 140 feet, if 
it is blowing at the rate of 12 knots? 24 knots? 50 knots? 



Water vapor is the most variable of the gases of the atmosphere, 
ranging from almost zero to a maximum of about four per cent by 
volume. It is extremely important to man's existence on the earth 
and constitutes one of the primary elements of weather. It not only 
contributes to the heating and cooling of the earth's surface but is 
directly related to the distribution and extent of precipitation over 
the earth. 

Humidity Observations 

Some of the molecules at the surface of a liquid are continually 
escaping and entering the air as gaseous molecules, thereby reduc- 
ing the volume of the liquid. For any surface of a given liquid, such 
as water, the number of molecules that escape in a period of time 
depends solely upon the speed at which they are moving, that is, 
upon the temperature of the surface of the liquid. Raising the tem- 
perature increases the velocity of the molecules and the rate at which 
they break free from the liquid surface. In this way, water vapor en- 
ters the air from water surfaces, moist soil, and growing plants. The 
process is called evaporation, and it occurs in all liquids. Ice and 
snow sometimes change directly from the solid to the gaseous state. 
This process is called sublimation. In breaking away from the attrac- 
tion of the other molecules, the escaping molecules use heat energy 
at the expense of the immediate environment. The heat energy so 
lost does not warm the gas but is used solely in effecting the change 
of state, and is called latent heat of vaporization, or latent heat of 



sublimation, as the case may be. This latent heat is again returned 
to the environment upon condensation of the water vapor. The wide 
use of evaporative cooling units throughout the western half of the 
United States is evidence that cooling by evaporation contributes 
much to the temperature characteristics of the atmosphere. 

Vapor pressure and saturated vapor. When water vapor escapes 
into space and mixes with the other gases of the air, it exerts a pres- 
sure in all directions, as do the other gases. This is known in mete- 
orology as the vapor pressure of the air. It is independent of the 
pressure of the other gases, exerting the same pressure when mixed 
with the other gases of the air as it would alone. The force exerted 
depends upon the concentration of the vapor, that is, upon the num- 
ber of molecules per unit volume. It is commonly expressed in the 
same units as the total air pressure, either in millibars or in inches 
or millimeters of mercury, referring to the length of the barometer 
column which the partial pressure due to the water vapor would 

Considering an open water surface, we find not only an escape of 
molecules from the liquid to the air, but also some return of the 
gaseous molecules to the liquid. At first the number escaping will 
be greater than the number returning, and we say that evaporation 
is occurring. But as the number of molecules of vapor in the air in- 
crease, there is an increase in the vapor pressure and in the number 
returning to the liquid, until a point is reached when the number re- 
turning is just equal to the number escaping. The net evaporation 
is then zero, and the air is said to be saturated; that is, the space can 
hold no more water vapor under the existing conditions. If we now 
raise the temperature of the air, the tendency of the water vapor to 
return to a liquid state is decreased, and we must add more vapor 
to keep the space saturated. At any given temperature, the satura- 
tion vapor pressure has a definite, fixed value, but the value changes 
rapidly with change of temperature, as may be seen in Table I, page 
48. For example: at 0F. the saturated vapor pressure is 0.038 inch; 
at 50F., 0.360 inch; and at 100F., L916 inches. These values have 
been determined by careful experiment. 

Dew point and condensation. The dew point of a given mass of 
air is the temperature at which saturation occurs when the air is 
cooled at constant pressure without the addition or removal of water 
vapor. The dew point is always expressed in degrees of temperature 


and is frequently compared with the temperature of the free air to 
determine humidity conditions. The dew point is determined by the 
vapor pressure of the air, however, and is entirely independent of 
the actual temperature. 

If the air is cooled below its dew point, some of the water vapor 
becomes liquid. This process of changing from a gas to a liquid is 
called condensation. As heat is transformed into work in the process 
of evaporation, resulting in cooling the liquid, so, in condensation, an 
equal amount of energy is transformed into heat, called latent heat 
of condensation, which results in adding heat to the air. Ordinarily 
condensation begins as soon as the dew point is passed; but under 
certain conditions condensation is delayed until the vapor is cooled 
considerably below its dew point. In that condition the air is said 
to be supersaturated. , 

Absolute humidity. In dealing with the moisture in the air, one 
quantity that may be measured is the actual mass of water vapor in 
a given sample of air. It may be expressed, for example, as the num- 
ber of grains' weight in a cubic foot of air, or the number of grams 
in a cubic centimeter. We thus obtain the absolute humidity, which 
is defined as the mass of water vapor per unit volume of air. 

Specific humidity. Another measure of humidity which has come 
into general use in meteorological studies, especially in connection 
with upper-air observations, is called specific humidity, defined as the 
weight of water vapor per unit weight of air ( including the water 
vapor). Notice that absolute humidity is the relation of the weight 
of vapor to the volume occupied, and specific humidity is the rela- 
tion of weight of vapor to weight of air. Again, since the pressures 
exerted by gases are proportional to their masses, specific humidity 
may be obtained by dividing the partial pressure due to the water 
vapor by the total pressure of the air. Thus we have the equations: 

Weight of vapor pressure of vapor ( e ) Ke 
Sp.Hum. (<jf) = 

Weight of air total air pressure ( p ) p 

where e and p are expressed in barometric units, such as millibars 
or inches, and K is a constant depending on the unit of specific 
humidity. Specific humidity is usually expressed in grams of water 
vapor per kilogram of air, and the equation becomes: q = 622 e/p. 
When a quantity of air expands or is compressed, the total pressure 


and the vapor pressure change in the same ratio, so that the value 
of e/p remains the same. Hence, the specific humidity is constant 
under these conditions; it does not change unless water is added or 

Mixing ratio. The mixing ratio is defined as the weight of water 
vapor per unit weight of completely dry air. It is the ratio of the 
water vapor to the remainder of the air. It thus differs from specific 
humidity only in using the pressure of the dry air instead of the total 
pressure of the air. Therefore, using w for mixing ratio, the equation 

in grams per kilogram. Since e is small as compared with p, the 
numerical values of mixing ratio differ little from those of specific 
humidity. * 

Relative humidity. Relative humidity is another manner of ex- 
pressing water-vapor content of the air. It may be defined as the ratio 
between the amount of water vapor present and the amount required 
for saturation under fixed temperature and pressure conditions. Rela- 
tive humidity is always expressed as a percentage or ratio. The 
amount of water vapor present in a unit volume of air is, by defini- 
tion, the absolute humidity; hence, the relative humidity equals the 
absolute humidity divided by what would be the absolute humidity 
if the air were saturated. Since vapor pressure, absolute humidity, 
specific humidity, and mixing ratio are different ways of expressing 
the same thing, relative humidity may be expressed in each case as 
the ratio between the existing condition and the saturated condi- 
tion. A common algebraic expression for relative humidity is: 

f = > 


where / is the relative humidity, e is the vapor pressure, and e is the 
saturation vapor pressure. 

Relative humidity and dew point are widely used by the layman 
to express water-vapor characteristics of the atmosphere, while 
vapor pressure, mixing ratio, absolute humidity, and specific humid- 
ity are commonly used only in scientific study of the atmosphere. 


Measurement of humidity. The dew point may be determined 
directly by a simple laboratory experiment. When water is placed in 
a thin-walled, brightly polished silver cup and kept well stirred, the 
temperature of the liquid and the cup will be the same. When suf- 
ficient ice is added to cool the water and the cup below the dew 
point of the surrounding air, the outer polished surface of the cup 
will be visibly clouded by beads of water. The temperature of the 
water (thoroughly stirred) at the time this clouding begins is the 
dew point of the surrounding air to a close approximation. An in- 
strument of this kind is called a dew-point hygrometer. 

The absolute humidity may be measured by passing a known 
volume of air through a chemical which absorbs the moisture, and 
noting the resulting increase in weight of the absorbing substance. 
By more elaborate instrumental means, the saturation vapor pres- 
sures at different temperatures have been experimentally determined 
with great care. From such laboratory determinations, and from the 
known physical relations between the various humidity factors, 
tables of humidity values have been prepared and published. The 
authoritative publications in this country are Smithsonian Meteoro- 
logical Tables, published by the Smithsonian Institution, and tables 
published by the United States Weather Bureau. 

Psychrometers. In meteorological practice, a psychrometer is 
commonly used for humidity measurements. The whirled psychrom- 
eter consists of two mercury thermometers with cylindrical bulbs, 
mounted vertically within the instrument shelter upon a frame that 
can be turned rapidly. The two thermometers are alike, but one has 
a thin piece of clean muslin tied around the bulb. This bulb is dipped 
in water, and the two are whirled. After a minute or two of whirl- 
ing, the two thermometers are read. The reading of the thermom- 
eter with the dry bulb is the current temperature of the air; the 
wet-bulb thermometer will ordinarily be found to have a lower read- 
ing. The whirling is repeated until no further reduction in the 
reading of the wet-bulb thermometer can be obtained. This read- 
ing is called the wet-bulb temperature. It remains constant as long 
as the covering remains wet and the whirling is continued, provided 
the air retains the same temperature and the same moisture content. 
The cooling of the mercury is due to the evaporation of the moisture 
around it and is directly proportional to the dryness of the air. The 
difference in temperature between the dry-bulb and the wet-bulb 



Fig. 22. Whirled Psyehrometer. For humidity observations, a \\ ( I inuslm slrrvr is 
fitted over the bulb of one of the thermometers, which are then vvhiilrd i.ipidly by 
turning the crank. Courtesy, Friez Instrument Division, Bendix Aviation Corp. 

thermometers, therefore, gives a measure of the moisture of the air. 
Given this difference, which is called the depression of the wet bulb, 
it is possible to read the dew point, vapor pressure, and relative hu- 
midity from the Smithsonian Meteorological Tables. A sling psy- 
chrometer is a similar instrument except that the two thermometers 
are mounted together on a metal back and are whirled by hand by 
means of an attached cord or chain. The aspiration psychromvter 
has the two thermometers enclosed in a tube through which air is 
drawn by a fan. 
A humidity-measuring device recently developed is called the 






tr o> 01 co 01 

I - co ^ io 


7 rS 






"* co o 3* 

00 O CO CD 

i^q<oi* i ;a5po5'' | i;CJ 




s i 

-3 CO 
W - 



CO 00 CO t * 



O 01 ^f JO 00 



telepsychrometer. As the name implies, it shows the wet-bulb and 
dry-bulb readings on an indicator dial mounted in the weather office 
while the measuring instrument is favorably exposed to the free at- 
mosphere some distance away. The telepsychrometer permits the 
observer, by means of electric controls, to perform the necessary 
operations for a reading without leaving the office. A heater element 
is incorporated to permit the telepsychrometer to function at tem- 
peratures below freezing. 

In Table I, condensed from Psychrometric Tables of the United 
States Weather Bureau, will be found the saturation vapor pressures 
for various temperatures, and also a table for obtaining dew points 
from psychrometric observations. Differences between the readings 
of the wet-bulb and dry-bulb thermometers are given at the head of 
the columns 1 to 30. The dew point corresponding to a given air 
temperature and a given depression of the wet-bulb thermometer is 
found in the body of the table. For example, when the temperature 
is 50 and the depression is 6, the dew point is 37; also, when the 
temperature is 65 and the depression 14, the dew point is 37. 
Similarly, Table II gives the relative humidity in terms of the air 
temperature and the cooling of the wet thermometer. It will be 
seen, for example, that the relative humidity is 55 per cent when 
the temperature is 20 and the depression 3 and again with a tem- 
perature of 70 and a depression of 10. 

Hygrometers. The hair hygrometer is an instrument which gives 
a direct reading of the relative humidity. The oils are removed from 
a strand of human hair, which is then attached so that its changes 
in length actuate a pen moving over a dial. In the hair hygrograph, 
the pen moves over a cylinder and makes a continuous record of 
the relative humidity. The hairs change their length in proportion 
to the changes in relative humidity, getting longer as the humidity 
increases. In using psychrometers or hygrometers, it must be re- 
membered that maintaining an active movement of air past the in- 
strument is essential to obtaining a correct reading. 

The psychrometer is rather inaccurate at temperatures below 
freezing, because ice forms on the bulb. The hair hygrometer needs 
frequent calibration and is slow in responding to humidity changes. 
The time lag increases as the temperature decreases, and is so great 
as to make the instrument practically useless at temperatures a lit- 
tle below zero F. These defects make the hair hygrometer unsatis- 


factory for use on airplanes or for other upper-air measurements be- 
cause of the low temperatures and rapid changes in humidity en- 
countered. An electric hygrometer has recently been developed to 
replace the hair hygrometer, especially on airplanes and radiosondes. 
It makes use of the fact that the resistance through an electrical con- 
ductor coated with a moisture-absorbing material varies as die rela- 
tive humidity varies. 

Humidity records. From readings of the psychrometer, one ob- 
tains the dew point, relative humidity, and vapor pressure. Many 
stations also obtain a continuous record of the relative humidity by 
means of the hair hygrograph (Fig. 23). The relative humidity 
shows both a diurnal and an annual variation; it is, on the average, 
greatest during the coolest part of the day and of the year, and least 
during the warmest portions. 

Vapor-pressure and dew-point data are used in many meteorolog- 
ical studies, in the forecasting of weather conditions, and in many 
practical applications. Variations in relative humidity have a direct 
effect on human comfort and health, as will be noted later, and also 
affect many of man's occupations. For example, maintenance of the 
correct relative humidity is important in keeping fruits, eggs, and 
other perishable products in good condition in cold storage. Such 
products as silks and cigars can best be manufactured where the hu- 
midity is rather high, but sun-dried fruits require a dry atmosphere. 

Evaporation Observations 

Evaporation is of primary interest in meteorology as the source of 
the water vapor of the air. It is also important in its effect on soil 
moisture and plant growth. 

Amount of evaporation. The depth of water evaporated from a 
given water surface in a given time depends in the main upon the 
following factors: 

1. Vapor pressure of the water surface. This is directly related 
to the temperature of the water surface. Increasing the temperature 
increases the vapor pressure and the evaporation, if other conditions 
remain unchanged. 

2. Vapor pressure of the air. The rate of evaporation varies di- 
rectly as the difference between the saturation vapor pressure at die 
temperature of the water surface and the existing vapor pressure in 









2 <o 





the air. The latter varies directly with the relative humidity of the 

3. Wind movement. Wind removes the moist air in direct con- 
tact with the water and replaces it with drier air. Hence, evapora- 
tion increases with wind velocity. 

4. Salinity. The presence of dissolved minerals or salts in the 
water retards evaporation. Evaporation from sea water is about five 
per cent less than from fresh water, other conditions being the same. 

Measurement of evaporation. The evaporation from a water sur- 
face is often measured by use of a shallow circular pan, 4 to 6 feet 
in diameter and 10 to 12 inches deep, This pan is filled with water 
nearly to the top and the decrease in depth is measured carefully 
every 24 hours by a hook gage (Fig, 24). The loss of water from 
such a pan will depend not only on the general factors mentioned 
in the preceding paragraph, but also on the size of the pan and its 
methods of exposure whether it is bur|ed in the ground, resting on 
the surface of the ground, or raised above ground with air circulat- 
ing beneath. All of these factors affect the temperature of the water. 
The evaporation from such a pan is not the same as from a lake 
under similar weather conditions, partly because the lake water 
takes on a different temperature, and partly because the moisture 
content of the air is increased in moving across a considerable body 
of water. 

Evaporation from plant and soil surfaces is great, and the rate is 
affected by other factors in addition to those applying to a water 
surface. In the case of the soil, evaporation is influenced by the 
texture and tilth of the soil and by its water content. In the case of 
plants, it varies for each species and, in the same species, with the 
leaf surface and the growing condition of the individual plant. No 
satisfactory formulas have been developed to connect the measured 
evaporation from a pan with the loss from larger bodies of water or 
from plants and soil, as the relations are complex in all cases. How- 
ever, records made in different localities with the same kind of pan, 
similarly exposed, give valuable comparative results, showing the 
relative amounts of evaporation in different climates. 

Another method of measuring evaporation has been developed by 
Thornthwaite and Holzman. 1 They mounted two small instrument 

i C. W. Thornthwaite and B. Holzman, "The Determination of Evaporation from 
Land and Water Surfaces," Monthly Weather Review, Vol. 67, 1939, pp. 4-11. 



Fig, 24. Class A Evaporation Station From left to right the instruments and equip- 
ment are an instrument shelter, weighing rain gage, evaporation pan with stillwell and 
anemometer, and a standard 8-inch rain gage. Courtesy. ('. S. Weather Bureau. 

shelters on a tower, one near the surface and one several feet di- 
rectly above the other. Recording instruments within these shelters 
give a continuous record of pressure, temperature, and relative hu- 
midity. From these data the specific humidities of the air at the two 
shelters, and the density of the air, are calculated. Continuous rec- 
ords of the wind velocity at the two points are also obtained. The 
differences between the specific humidities and between the wind 
speeds at the upper and lower points, taken together with the density 
of the air, give a measure of the vertical flow of water vapor. That 
is, they determine the mass of water moving upward in a given time, 
and hence the loss by evaporation from the surface, including the 
transpiration from plants. The method can be used over either land 
or water surfaces. 

In the greater part of the western half of the United States, where 
precipitation is light, the annual evaporation from a water surface 
is greater than the annual rainfall. In parts of Ari/ona it has been 
found to he more than nine times the laiuiall In spite of difficul- 
ties in the application of evaporation data to specific problems, such 
data are of much practical value; for example, to hydraulic en- 


gineers, in the planning of storage reservoirs and irrigation systems; 
and to plant scientists, in the study of the relations of plants to their 

Cloud Observations 

Clouds are condensed moisture, consisting of droplets of water or 
crystals of ice, having diameters varying from 0.001 to 0.004 inch. 
They are easily sustained and transported by air movements as slow 
as one-tenth of a mile per hour. 

Cloud classification. Although clouds are prominent and often 
spectacular features of the sky in nearly all parts of the world, there 
seems to have been no attempt to name and classify them until an 
Englishman, Luke Howard, in 1803, suggested the classification 
which has become the basis of all later cloud nomenclature. How- 
ard's system was based on the appearance of the clouds to the ob- 
server on the earth, but has been mocjlified from time to time to 
conform to increasing knowledge of the physical processes in the 
formation of clouds of different types. The classification now in use 
throughout the world is known as the international classification 
and is sponsored by the International Meteorological Committee, 
representing the official meteorological services of the world. It was 
first published in 1896 and has been revised at intervals. The latest 
revision defines and describes ten types of clouds, retaining the 
three main type forms: cirrus, stratus, and cumulus, first named by 
Howard. The other names are derived by combinations of these 
three words and by the use of alto, meaning high, and nimbus, mean- 
ing rain cloud (Fig. 25). In addition, there are numerous subtypes 
or varieties. 

International cloud forms. The ten forms now given official in- 
ternational names and descriptions are as follows ( the abbreviations 
are also in international use) : 

1. Cirrus (Ci. ). Detached clouds of delicate and fibrous appear- 
ance, without shading, generally white in color, often of a silky ap- 
pearance, of varied forms, such as tufts and featherlike plumes, and 
often arranged in bands. Clouds of the cirrus family are composed 
of ice crystals. Light rays from the sun or the moon frequently cause 
a halo to be visible. Because of their great altitude, these clouds 
often reflect beautiful hues of red or yellow before sunrise and after 


2. Clrrocumulus (Cc.). A group of small white flakes or very 
small globular masses, without shadows, associated with cirrus or 
cirrostratus, often arranged in rows; rather rare. 

3. Cirrostratus (Cs.). A thin, whitish veil which does not blur 
the outline of the sun or moon but gives rise to halos; sometimes it 
merely gives the sky a milky look, sometimes shows a fibrous struc- 
ture with disordered filaments. 

4 Altocumulus ( Ac. ) . A layer or patches composed of flattened, 
globular masses with or without shadows. The globules frequently 
have definite dark shading. Sometimes they occur in a regular pat- 
tern or lines or waves, producing what is called a mackerel sky. Some 
varieties are closely packed, approaching altostratus, and some have 
a vertical development, suggesting cumulus clouds. 

5. Altostratus (As.). Striated or fibrous veil of ground-glass ap- 
pearance, more or less gray or bluish in color, like thick cirrostratus 
but without halo phenomena; the sun or moon shows vaguely or is 
completely hidden. Rain or snow may fall from altostratus, and 
from any of the following forms. 

6. Stratocumulus (Sc.). A layer or patches of flakes or globular 
masses; the smallest of the regularly arranged elements are fairly 
large; they are soft and gray with darker parts, arranged in groups, 
lines, or rolls. Often the rolls are so close that their edges join 
together; when they cover the whole sky, they have a wavy appear- 

7. Stratus (St.). A uniform layer of cloud, resembling fog, but 
not resting on the ground. When this low layer is broken up into 
shreds, it is designated fractosstratus (Fs.). 

8. Nimbostratus (Ns. ). A low, amorphous, rainy layer of dark- 
gray color and nearly uniform. When it gives precipitation, con- 
tinuous rain or snow results; but it should l>e culled nimbostratus 
even when no rain or snow falls. 

9. Cumulus (Cu.). Thick clouds with vertical development; the 
upper surface is dome-shaped and exhibits protuberances, while the 
base is nearly horizontal. When the light comes from the side, the 
clouds exhibit strong contrasts of light and shade; against the sun, 
they look dark with a bright edge. A broken cloud resembling a 
ragged cumulus in which the different parts show constant change 
and are without well-defined upper and lower limits is called frac- 
tocumulus (Fc.)' Two distinct types of cumulus are recognized. 


Cumulus humilus, or cumulus of fair weather, have limited vertical 
development. Bases and tops are clearly defined and tend to persist 
in a fairly constant state. Cumulus congestus, or towering cumulus, 
shows marked vertical development. They may have well-defined 
bases, but their tops often "boil," showing evidence of atmospheric 
instability and strong vertical air currents. Cumulus congestus may 
develop into cumulonimbus. 

10. Cumulonimbus (Cb.). Heavy masses of cloud with great 
vertical development, whose summits rise in the form of mountains 
or towers, the upper parts having a fibrous texture and often spread* 
ing out in the shape of an anvil; generally producing showers of rain, 
snow, or hail. The base often has a layer of low, ragged, fractostratus 
or fractocumulus clouds below it. Masses of cumulus clouds, how- 
ever heavy they may be, should not be classed as cumulonimbi un- 
less the whole or a part of their tops is transformed, or is in process 
of transformation into a cirrus mass. * 

Height and grouping of cloud forms. The elevation of clouds 
varies greatly among the different types and also in the same type 
on different occasions and in different latitudes. The ten forms are 
classified into four groups, or families, as follows: 

High clouds. Cirrus, cirrocumulus, cirrostratus; average height, 
4-7 miles (6-11 km). 

Middle clouds. Altocumulus, altostratus; average height, 1-4 
miles (2-6 km). 

Low clouds. Stratocumulus, stratus, nimbostratus; average 
height, 300 to 6500 feet (0.1-2 km). 

Clouds with vertical development. Cumulus, cumulonimbus; 
average height, 1600 feet at base to 4 miles at top (0.5-6 km). 

The three cirrus forms and the cirriform tops of cumulonimbus 
are composed of ice crystals; the other clouds are made up of water 
droplets but may contain some ice crystals. 

Cirrus, cirrocumulus, altocumulus, and cumulus ckmds occur in 
detached masses, usually covering only a part of the sky, and may be 
called fair weather clouds, since rain normally does not fall from 
these forms. The remaining types, cirrostratus, altostratus, Strato- 
cumulus, stratus, nimbostratus, and cumulonimbus, form more or 
less continuous layers and often cover the entire sky. Precipitation 
may occur from any of these except cirrostratus. Cumulus and cumu- 
lonimbus are of great vertical depth, the tops of cumulonimbus 



ilf t 


* let 

its suit 




!*! or 

Fig* 1 A of 

A, j, trot/ 



Fig. 26. Cimis. Delicate cirrus, composed of irregularly arranged filaments oriented 
In various directions; showing at lower left a tendency to fuse together into cir- 
rostratus. Photo by H F/orrrn. 

Fig. 27. Cirrocumulus. Closely packed, small globular masses or flakes, arranged 
in lines or ripples; associated wfth cirroftratuj at lower right. U. S. Army, Photo 
Section, Lake Charles, La. t Flying School. 



Kitf 28. Cirrostratns. A thin veil at uppei left, and fibnms structure with parallel 
hands at left center, patches ot altocumulus and lenticular altostrutus below. Photo 
by //. f'/um n 

Fig. iJ, Altocumulus. A layer of altocumulus at rme level compustxl of v>ft, ilat 
rt>umlixl masses with a stmcture in two diroctn>jis, thick enough to be rather heavily 
shaded in places, hut with interstices where the blue sky appears. Photo by H. b lore en. 


Fig. 30. Altocumulus in Patches, with Stratoeumulus Below and Altosh.itus ,t Upper 
Right. Plwto by //. Moreen. 

Fig. 31. Altocumulus. Bands of altocumulus advancing from left to right across the 
sky. Couri&tg, U. S. Weather Bureau and Willard Wood. 



Fig. 32. Altostratus Veil covering Entire Sky. Two well-defined masses of lenticular 
altocumulus are at the center. P/u;fo by H. Floreen. 

Fig. 33. Altostratus. Well defined altostratus layer above stratocuimiius and fog in 
the valley below. Courtesy U. S, Weather Bureau. 



Fig. 34. Stnitocumuliis Layer. Light and shade contrasts are apparent near the /emth 
and a roll structure near the hori/on Photo by II. Moreen* 

Fig. 35. Stratocumulus in a Broken Layer. Masses are arranged in rolls and lines; 
dark shadows indicate considerable thickness of the clouds. Courtesy, U. S. Weather 
Bureau and Ansco. 



|JT ", /AVV,, /' ^ ,K ," r ; L 

l'i. -to Stiutus Low stratus < InuiK mmum uv.t a small, rocky island. These 
clouds most; in at very low altitudes ovri lla- \\rst C-oast to form fog. Courtesy, U. S. 
Weather Bureau, A#fa and Ansco. 

Fig. 37. Ninvtxxstr.Uus llu \\i\At i 
of a cold front. Liht MHU \\ is t.dliu 
not shown and the wsibihu i> shm 
Wotffarr Bureau and Jo/in C>. Ward. 

>!T of nnulx^ti.itns dmid* during the 

R hut because of the use ol infra-red film ft is 

greater than it actually was. Courtesy, V. S. 


*-" - 

Fig, 38. Cumulus of Finr Weather S< altered in, ISM s \\ilh a Hat and deflated ap 
pearance, the horizontal extension Ix-mi; reatei than tin <>nl\ hi'ht s 
are apparent showing that the thirkness is iut K r< '' 1 /'/*"/< />!/ // H<rri 

Fig. 39. Cutmtlonimbus. CiunnlonimfnB cloud with bulging top thon^li without 
dence of cfnoform parts. Caurtety, V. S. Weather Bureau and It / lrcen. 



Fig. 40. Oumulomml>us with Anvil Top. An aerial view of the cumulonimbus anvil 
which is fiaxol out into cirrus forms having a structure very different from the 
rounded tumulus forms below. Photo by H. Floreen. 

of ton extending 2 to 5 miles (3-8 km) above their bases. The other 
1 01 ins are more like la\ns or sheets, comparatively thin but of great 
hon/onlal extent. These general characteristics help to identify the 
\arious loims, hut it should be remembered that there are all man- 
ner of giadations between them, and in some cases one form merges 
impereeptihU into another. Frequently it is not possible to identify 
clouds with eeitamt>\ unless one has watched their evolution or can 
interpret the physical processes that are producing them. 


Records of clouds and cloudiness. When a weather observation 
is being made, the kinds of clouds visible should be recorded, the 
amount of each kind, and the direction from which each is moving. 
The amount is estimated in tenths of the sky covered, Such observa- 
tions give information with regard to the direction, velocity, and 
turbulence of the wind at different elevations, and are often of direct 
aid in foreseeing weather changes. The Weather Bureau records 
each day as clear, partly cloudy, or cloudy, according to the aver- 
age cloudiness during the time between sunrise and sunset; clear, if 
the average cloudiness is three-tenths or less; partly cloudy, if be- 
tween four- and seven-tenths; and cloudy, if eight-tenths or more, 
Sky conditions at the time of making an observation are reported in 
greater detail, as described later. 

Precipitation Observations 

Precipitation, in meteorology, means either the falling of moisture 
to the earth in any form, or the quantity of water so deposited, ex- 
pressed in depth of water. Precipitation takes various forms, such as 
rain, snow, hail, and other special formations, which are discussed 
in more detail in the chapter on condensation; but dew, frost, and 
fog are not regarded as precipitation. Amount of precipitation al- 
ways means the liquid content. The word rainfall is often used as 
synonymous with precipitation, meaning the amount of water in 
whatever form it may have fallen. 

Rain gages. The object of a rainfall measurement is to obtain 
the thickness of the layer of water that has fallen, assuming it to be 
evenly distributed over the surface in the vicinity of the measure- 
ment. Any open vessel of the same cross section throughout and ex- 
posed vertically will serve as a rain gage. A cylindrical vessel is 
preferable. The accuracy of measurement can be increased by meas- 
uring the catch in a vessel smaller than that in which it is received, 
if the ratio of the cross sections of the two vessels is accurately 

The 8-inch rain gage. The 8-inch rain gage is a cylindrical re- 
ceiver exactly 8 inches in diameter, provided with a funnel-shaped 
bottom (Fig. 41). This funnel conducts the rain caught by the re- 



ceiver into a cylindrical meas- 
uring tube 20 inches long and 
one-tenth of the cross-sectional 
area of the receiver. The depth 
of the rainfall is, accordingly, 
magnified just 10 times. For 1 
inch of rain, the water is 10 
inches deep in the measuring 
tube. Thus, the amounts can 
easily be measured with great 
precision. The depth is meas- 
ured by a small rule or measur- 
ing stick, graduated in inches 
and tenths. The receiving fun- 
nel fits over an outer tube, 
8 inches in diameter, which 
serves to support the receiving 
funnel and also to hold excess 
water when more than 2 inches 
of rain falls, as the inner meas- 
uring tube, which is 20 inches 
long, consequently overflows. 
The funnel and the outer tube 
protect the water caught in the 
inner tube from appreciable 

Recording gages. The tipping-bucket recording rain gage has a 
10-inch receiving funnel, at the mouth of which is a bucket of two 
compartments, so mounted that one or the other receives the water 
coming from the funnel. As one compartment fills, it tips, thereby 
emptying its water and presenting the other compartment to the 
mouth of the funnel. The compartments are of such size with ref- 
erence to the receiving funnel that each tip represents 0.01 inch. 
As the bucket tips, it closes an electric circuit connected with the 
meteorograph. A permanent record is thus obtained of the time of 
occurrence of each 0.01 inch of rain, and consequently the amount 
of the fall in any given period. 

The Fergusson weighing rain and snow gage is also in common 
use. In this instrument the accumulating weight of the water or 

Fig. 41. Rain Gage. Standard 8-inch 
pattern, with measuring stick. Courtesy, 
Frlez Instrument Division, Bendix Avia- 
tion Corp. 



snow caught in an 8-inch cylinder moves a platform supported by a 
spring balance. This movement is communicated to a pen, which 
then writes a continuous record on a drum driven by a clock (Fig. 
42). The record reads directly in inches of precipitation rather than 
in units of weight. It indicates the rate of fall for any desired interval 
in addition to total fall since the previous reading. There are other 
devices for obtaining continuous records of rainfall, one of which 
is a method of recording the movement of a float that rises as water 
accumulates in the gage. 

Exposure of rain gage. To 
obtain a correct catch of rain 

with any gage, the gage should ^:^^/^? ; '!''' " ! 

be exposed on the ground in 
an open, level space at least 
as far from trees, buildings, or 
other high objects as they are 
high, in order that rain fall- 
ing obliquely may not be inter- 
cepted. Windbreaks at a dis- 
tance greater than their height 
are desirable as a means of 
checking the velocity of the 
wind. If the gage is placed 
much above the ground, the 
higher wind velocities carry 
more of the water around and 
over it. An exposure on the 
edge of a roof is especially bad 
because of eddies of wind 
around the gage in that loca- 

Measurement of snow. Two quantities are desired in the meas- 
urement of snowfall, namely, the actual depth of the snow and its 
water equivalent. Both measurements present some difficulty in 
practice. Snow does not ordinarily lie at a uniform depth over the 
ground, but drifts, even in moderate winds. Hence, measurements 
of the depth should be made at several different places, apparently 
representative, and the average of these measurements taken as the 
depth of the snow and recorded in inches and tenths. 

Fig. 42. Fcrgusson Weighing Rain Gage. 
Outside cover has been removed; makes a 
permanent, continuous record of the rate 
and amount of rainfall. Courtesy, U. S. 
Weather Bureau. 


To obtain the amount of precipitation from snow, that is, its water 
equivalent, the snow may be melted and measured as rain, or it may 
be weighed. Since snow readily blows around the top of a gage in- 
stead of falling into it, the amount caught by the gage is ordinarily 
considerably less than the actual fall. Gages with windshields around 
the cylinder, intended to break up the eddies and insure a more 
nearly correct catch, have been devised and are widely used, espe- 
cially in mountain regions. To obtain a representative sample for 
melting, the outer, overflow tube of the 8-inch gage may be forced 
downward through a layer of snow, representing the average depth 
of fall, and the section thus cut out may be lifted up by placing a 
thin board or sheet of metal underneath it. This sample should be 
melted and poured into the measuring tube and the depth deter- 
mined as in the case of rain. The best way to melt snow without 
loss is to add a measured quantity of warm water to it. The Fergus- 
son weighing gage may be used for snow as for rain, but the amount 
caught is subject to error. 

The amount of water in a given volume of snow varies greatly, ac- 
cording to the texture of the snow and the closeness with which it 
is packed. The texture of snow changes with its temperature from 
dry and feathery to moist. The closeness of packing depends not 
only on the texture as it falls but also on the depth of fall, the length 
of time it has lain on the ground, and the temperatures to which it 
has been subjected since falling. In moist, newly fallen snow, 6 
inches of snow may make 1 inch of water, while in small amounts of 
dry, fluffy snow, the ratio may be as high as 30 to 1. In the absence 
of definite information, the ratio of 10 inches of snow to 1 inch of 
water is frequently used as an average. In large drifts which have 
accumulated all winter and are melting in the spring, 2 inches of 
snow may be equivalent to 1 inch of water. 

Precipitation records. A daily precipitation record should include 
the kind and amount of precipitation, and the time of beginning 
and ending. A record is made of the depth of snowfall since the last 
observation and also of the total depth on the ground. From the re- 
cording rain gage, the amounts of rain in 5, 10, 15, and 30 minutes 
and in 1, 2, and 24 hours should be tabulated. When excessive rain- 
fall occurs, the accumulated amounts in successive 5- or 10-mimite 
periods may be tabulated. Rainfall is considered excessive when it 


falls at the rate of 0.25 inch in 5 minutes, 1 inch in 1 hour, or 2*50 
inches in 24 hours. 

From a long series of such records at a fixed point, normal daily, 
monthly, and yearly values may be determined; also, the greatest 
and least yearly and monthly amounts, and the greatest amounts in 
short periods of from 5 minutes to 48 hours, The record should also 
show the number of rainy days by months and by years, and their 
averages over the period of record. A rainy day is a day on which 
0.01 inch or more of precipitation occurs; that is, a day on which 
only traces of rain fall is not counted as a rainy day in this country. 
Unusually long periods without rain, known as droughts, should also 
be noted because of their great economic significance. 

Sunshine Observations 

The sun is the source of energy that evaporates water and causes 
wind to blow. By its direct rays, it causes some regions of the earth 
to be hot, and the lack of solar radiation leaves some regions very 
cold. Thus the sun is directly or indirectly responsible for the dis- 
tribution of rainfall and temperature over the earth. Nearly all plant 
and animal life require some exposure to sunlight if they are to en- 
joy good health or even to survive. Sunshine, therefore, is an impor- 
tant element of the weather. 

Sunshine recorders. The electric sunshine recorder, or sunshine 
switch, in common use has the form of a straight glass tube with cy- 
lindrical bulbs at each end, one bulb being smoothly coated on the 
outside with lampblack (Fig. 43). The air in the bulbs is separated 
by a small quantity of mercury and alcohol. This tube is protected 
from the influence of the air temperature by being sealed within an 
outer glass tube, the space between the two tubes being exhausted of 
air. Two wires with ends a short distance apart are sealed in the inner 
tube midway between the bulbs and are connected with the mete- 
orograph, which, as we have seen, also records wind direction, wind 
velocity, and rainfall. When the sun shines on this instrument, the 
black bulb absorbs radiant energy, and the mercury is warmed and 
expands to surround the ends of the two tubes, permitting an electric 
current to pass from one to the other. The circuit passes through the 
meteorograph, where it is closed each minute by a rotating contact 



point on the clock. The completing of the circuit results in a move- 
ment of the pen on the cylinder. A mark on the register sheet is thus 
made each minute when the sun is shining. When the sun is not 
shining, the bulb cools by radiation, the mercury drops below the 
contact point, the circuit is broken, and the pen moves across the 
sheet in a straight line. 

Fig. 43. Electrical Sunshine KecoultT. Courtesy, Friez Instrument Division, Bendix 

Aviation Corp. 

The use of this instalment depends upon the fact that lampblack 
becomes warmer than clear glass, under direct sunshine. The in- 
strument gives a record of the number of minutes of sunshine, with 
some lag at the beginning and ending of the sunshine, but no rec- 
ord of the varying intensity of the sun's rays. It is usually adjusted 


to record whenever the sun is bright enough to cast a visible shadow, 
except for the half -hour or more after sunrise and before sunset. At 
these times, the sun's rays are too weak, that is, are received at the 
instrument with too little energy, to operate it. The Campbell-Stokes 
recorder, used in many other countries and at some stations in the 
United States, consists of a glass globe which focuses the sun's rays 
and thus chars a track on a graduated card. The sun itself thus 
writes with fair accuracy an original record of the duration of sun- 

Sunshine records. From such instruments records are obtained 
of the duration of sunshine each day. The percentage obtained by 
dividing the actual sunshine by the possible sunshine for the day is 
usually calculated and recorded. For the month and the year, the 
total number of hours of sunshine and the percentage of the possible 
are important data. 

Observations of Visibility and Ceiling 

The advent of air navigation at great speed has given increased 
importance to the clearness of the atmosphere as a weather element, 
meaning the distance at which objects can be seen, and the height 
of the clouds. Information on these matters is obtained by observ- 
ing visibility and height of ceiling. 

Visibility. As officially defined for purposes of observation and 
record, visibility is the greatest distance toward the horizon at which 
prominent objects, such as mountains, buildings, towers, lights, and 
so forth, can be seen and identified by the unaided eye. This dis- 
tance depends upon the clearness of the air, and is modified by the 
turbulence of the air and by the presence of haze, dust, smoke, fog, 
rain, and snow. Visibility records are made by eye observations of 
stationary objects at known distances from the observer, and may be 
expressed either in miles or in fractions of miles, or indirectly ac- 
cording to an arbitrary scale of numbers. 

The problem of measuring visibility is complex, for several rea- 
sons. First, there is a human error possible because all people do not 
see alike. Also, objects vary in their ability to cast a distinct visible 
outline, a fact which leads to errors in estimation. The distance at 
which an object can be seen depends in part on its color and to a 
greater degree on the direction of lighting. Finally, visibility is meas- 



ured horizontally, while haze or other obstruction to vision may 
affect vertical visibility in a very different way. 

Ceiling. The word ceiling came into meteorological use, with the 
development of aviation, to designate the height of the base of an 
extensive cloud layer that prevents an observer on the ground from 
seeing above the layer and prevents the pilot aloft from seeing the 
ground. It is always expressed in feet above the ground. For pur- 
poses of observation and record at airways stations, "ceiling" is now 
defined as the height ascribed to the lowest layer of clouds or other 
obscuration phenomenon that is classed as broken, overcast, or obscu- 
ration and not classed as "thin" or "partial." In the case of obscura- 
tion phenomenon, such as blowing dust, this height represents actual 
vertical visibility entering the obscuration rather than the base of 
the phenomenon. Ceiling is unlimited at all other times. 

The height of the ceiling may be determined by day by the use 
of ceiling balloons. These are small rubber balloons inflated to rise 
at a definite rate. This rate multiplied by the time elapsed between 
release of the balloon and its disappearance in the cloud equals the 
height of the base of the cloud. The ceiling balloon may be used at 
night also, if a small light is attached to it. More common, at night, 
is the use of a ceiling light. A beam of light is projected vertically 
on the base of the clouds, and the height is measured trigonometri- 
cally by measuring the angular elevation of the top of the beam 
along a known base line ( Fig. 44 ) . 

1000 FEET 

Fig. 44. Measuring Ceiling Height with a Ceiling Light. Drawing by Cory Steicart. 


The better-equipped airports of the United States began using the 
ceilometer as soon as it became available after World War II. This 
instrument uses the triangulation principle of the ceiling light, but is 
completely automatic in operation. The height of the vertical beam 
is determined by a very sensitive photoelectric scanner. Ceilings up 
to 10,000 feet ( 3 km ) may be accurately measured by this method, 
either day or night. A timed graphing device is attached to give a 
continuous record of ceiling heights at the station. 

Upper-Air Observations 

The conditions existing in the free air above the earth's surface 
are important to theoretical studies of the atmosphere and also form 
one of the basic tools for weather forecasting. The development of 
aviation has greatly accelerated the search for knowledge of the 
upper atmosphere. * 

Obtaining upper-air data. In Europe and America, mountain 
observations have been made regularly since 1870. These observa- 
tions have given valuable information, but the conditions on moun- 
taintops are not representative of free-air conditions at the same 
altitudes. At various times during the nineteenth and twentieth cen- 
turies, free or captive balloons manned by one or more passengers 
and equipped with meteorological instruments were released to ex- 
plore the upper air. These flights produced interesting and valuable 
data, but it became evident that a more complete and regular cov- 
erage of upper-air conditions was desirable. Free balloons bearing 
recording instruments and an offer of reward to the finder for their 
return also proved less than satisfactory. 

From 1898 to 1933, the United States Weather Bureau maintained 
kite-flying stations at which systematic upper-air records were ob- 
tained. Box kites carrying light recording instruments sometimes 
reached a height of four miles, but no flight could be made when 
the surface wind was not strong enough to launch the kite. This 
operation was replaced in 1933 by regular airplane flights from sev- 
eral points over the United States. These flights were synchronized 
and ascended over their respective stations as nearly vertically as 
practical to an elevation of about 17,000 feet (5.2 km). Such flights 
were expensive and required about \% hours for completion, and 
the records were not available until the flight was completed. Since 


1938, airplanes have been replaced by radiosondes in the regular 
network of upper-air observations in the United States. In addition 
to the radiosonde, other types of upper-air data are currently ob- 
tained by pilot balloons, rawins, aircraft reconnaissance, and rocket 
research. These methods will be discussed in greater detail. 

Pilot balloons. When fully inflated with helium or hydrogen, 
the pilot balloon is about 30 inches in diameter. It floats freely in the 
air but carries no instruments. The course of the balloon is watched 
and plotted from the ground by the use of a telescopic theodolite. If 
two theodolites at a known distance apart are used, the height and 
position of the balloon at the end of each minute can be computed 
accurately. If only one theodolite is used, it is necessary to assume 
a certain rate of ascent, and for this an empirical formula is used. 
From such observations, the direction and speed of the wind at vari- 
ous levels and the height of the clouds can be quickly determined. 
Regular flights of pilot balloons every 6 hours are made at a large 
number of weather stations throughout the United States, particu- 
larly along and near the main air routes. The soundings thus ob- 
tained are called pibals. 

Radiosondes. Since 1938, the use of radiosondes throughout the 
world has been increasing rapidly. The device consists of a small box 
containing temperature, humidity, and pressure instruments and a 
miniature radio-sending station. It is carried aloft by a large gas- 
filled balloon which is also equipped with a parachute to lower the 
instruments harmlessly to the ground after the balloon bursts ( Fig. 
45). The balloons are of good quality and frequently rise to a height 
of 15 or 20 miles (24-32 km) before bursting. In a systematic fash- 
ion, the three weather elements are measured and transmitted by 
radio to a receiving station on the ground. The records become avail- 
able for immediate use while the radiosonde is still in flight. Eleva- 
tions corresponding to the reported pressure levels can be computed 
very accurately when the temperature and humidity characteristics 
of the air column are known. Records from the radiosonde are 
known as raobs. They provide some of the most extensive and reli- 
able data available on upper-air conditions. Severe thunderstorms 
and heavy rains may cause instrument failure or interfere with radio 
reception, but otherwise the radiosonde may be used in all lands of 



Fig. 45. Radiosonde in Flight. Courtesy, Friez Instrument Division, Bendix Aviation 



Rawinsondes. Although the pilot balloon is an inexpensive and 
fairly accurate source of upper-air wind data, it disappears from 
sight when there are clouds, or it soon drifts beyond the range of 
the theodolite with strong winds. The development of radar and 
radio-directional receiving techniques during World War II made 
the rawinsonde (rawin) possible. This instrument consists of a track- 
ing device at the station which measures the direction and angular 
elevation of an ascending balloon. It is usually used with the radio- 
sonde and no extra balloon is necessary. By combining this observa- 
tion with the radiosonde record, wind direction and speed at all de- 
sired elevations or pressure levels can be easily determined. About 
200 radiosonde and rawin stations are operated in the United States 
by the Weather Bureau, the Air Force, and the Navy. 

Aircraft weather reconnaissance. With the development and ex- 
pansion of aviation and the increased need for complete weather 
information in the military operation of aircraft, the use of airplanes 
in the collection of upper-air data has been resumed and expanded 
under the name of weather reconnaissance. Airplanes are dispatched 
to learn of sky conditions over a selected target and along the route 
to the target, or to keep watch on the weather over a considerable 
area and make radio reports of existing conditions. As a part of the 
peacetime activities of the United States Air Force, flights by air- 
craft carrying meteorologists and specially designed meteorological 
instruments are scheduled at regular intervals along chosen patterns 
for the purpose of obtaining complete records of atmospheric con- 
ditions in areas where raobs and other vipper-air reports are missing. 
In other words, the airplane becomes a flying weather station that 
covers not only a large area, but a changing area that may be chosen 
on each flight as most likely to provide important information. Radio- 
sondes are sometimes dropped from these planes, transmitting re- 
ports of weather conditions as they descend by parachute. Such in- 
struments are called dropsondes. 

One of the most important and exciting aspects of aircraft weather 
reconnaissance is locating and tracking hurricanes and typhoons. 
These severe tropical cyclones originate over the oceans, in regions 
of little data at low latitudes, and have caused great property dam- 
age and loss of life upon arriving unannounced at a populated coast- 
line. Aircraft reconnaissance has become the principal source of 
weather data useful in reducing the destruction by these storms. 


Planes fly around and into the storm centers to determine their exact 
location, size, and intensity. Although progress has been made to- 
ward forecasting tropical cyclones, the surest way to avoid unneces- 
sary losses to their fury is to examine them at regular intervals by 
reconnaissance aircraft. 

Rocket-sonde research. A very new aspect of upper-air research 
has been introduced through the use of rocket-borne instruments. 
Most of this work has been accomplished under government spon- 
sorship since World War II at White Sands, New Mexico. 2 Rockets 
opened new possibilities for exploring the upper air, for, where bal- 
loons have on a few occasions reached a maximum height of about 
25 miles (40 km), some rockets have reached 10 times that figure. 
As a result of rocket research, the scientists' concept of temperature 
and wind distribution at high altitudes has been radically revised. 
These data are especially valuable to the theoretical meteorologist, 
and they may also be shown to have direct bearing on the weather 
at the earth's surface. 

A standard weather observation. In making a complete weather 
observation, the observer determines and records the following 
values and conditions: 

Clouds. Amount, kind, and direction of movement, as estimated 
by the observer from eye observation. Cloud height is given in 
hundreds of feet, estimated or obtained by ceiling balloon or ceilom- 

Ceiling. Expressed to hundreds of feet. 

Visibility. By eye observation, expressed in miles or fractions. 

Present weather. Clear, partly cloudy, cloudy, overcast, raining, 
snowing, and various special conditions. 

Past weather. Weather conditions since previous observation, 

Dry-bulb thermometer. Air temperature at time of observation. 

Wet-bulb thermometer. Wet-bulb temperature at time of obser- 
vation; gives means of calculating dew point, relative humidity, and 
vapor pressure. 

Maximum thermometer. Highest temperature since previous ob- 

Minimum thermometer. Lowest temperature since previous ob- 

2 Homer E. Newell, Jr., High Altitude Research. New York; Academic Frew Inc., 
1953, pp. v-vii. 


Precipitation. Amount since previous observation. If in the form 
of rain, the amount is determined by stick measurement of the water 
in the measuring tube of the rain gage and checked with the rec- 
ord on the register sheet. Depth and water equivalent of snow are 
measured, as has been explained. The character of all precipitation 
and the times of its beginnings and endings are also recorded. 

Wind direction. Obtained by observing the wind vane; con- 
firmed by reference to triple register or wind indicator. 

Wind velocity. From indicating or recording anemometers. 

Maximum velocity. For a five-minute period since last observa- 
tion; from recording anemometer. 

Station pressure. Barometer reading corrected for temperature. 

Reduced pressure. Station pressure reduced to sea level by cor- 
rection for altitude. 

Pressure tendency and change. During the past three hours; 
from barograph record. 

All the observed and calculated values should be entered on a 
large form, which constitutes a permanent original record of the 
weather. They are the fundamental data for studies of weather and 
climate, and they have many specific, practical applications. When 
simultaneous observations from many stations are collected by tele- 
type and the data entered on an outline map, a daily picture of 
weather conditions is obtained. Such a map is called a synoptic 
weather map and is a principal basis of weather forecasting. 


1. Find dew point, relative humidity, and vapor pressure when the 
dry bulb thermometer reads 60 and the wet bulb thermometer reads 52. 

2. Find dew point and vapor pressure when the temperature is 73 
and the relative humidity 28 per cent. 

3. Find relative humidity and vapor pressure when the temperature 
is 35 and the dew point 21. 

4* Outside air at a temperature of 25 and a relative humidity of 62 
per cent is taken into a room and warmed to a temperature of 75 with- 
out the addition or loss of moisture. What is its new relative humidity? 

5* Air having a temperature of 50 and a relative humidity of 49 per 
cent cools during the night to a temperature of 32. What does its rela- 
tive humidity become? 

6, If the total pressure of the air at a given time and place is 29.620 


inches, and the partial pressure due to the water vapor is 0.250 inch, what 
is the specific humidity of the air in grams per kilogram? 

7. A fall of one inch of rain amounts to how many tons of water per 
acre? Per square mile? ( A cubic foot of water weighs 62.4 pounds, ) 

8. The area of North Carolina is 54,426 square miles, and its average 
July rainfall is 5.78 inches. What is the weight of the water that falls? 

9. What is the diameter of the cylindrical measuring tube used with 
the 8-inch rain gage? 

10. Make a list of all the ways that direction and speed of winds above 
the ground may be determined. If you knew the height of a cloud, how 
could you measure the wind velocity at cloud level? 



In observing and measuring the weather and its changes, we have 
noted how extraordinarily variable is the temperature of the air. 
It is now necessary to examine more closely the way the air is heated 
and cooled, and into some of the physical effects of its variations in 
temperature, for most of the phenomena of the weather have their 
origin in temperature changes. 

Radiant Energy 

If you stand before a fireplace, the heat that reaches you from the 
burning coals is said to travel through the intervening space as radi- 
ant energy. It would reach you in the same way if there were no 
air in the space. The fuel loses the energy which is thus sent out 
through space in a form having many of the characteristics of trans- 
verse waves. These are known as electromagnetic waves, and the 
energy thus transferred is called radiant energy or radiation. 

Radiation. Radiation refers both to the radial emission of energy 
from an object and to the energy so transferred. The movement of 
energy through "empty space" in a manner suggesting waves, but 
without the agency of any material medium, is mysterious. But 
there is complete evidence that waves of energy do travel in this way 
and that every object in the universe, whether hot or cold, has the 
faculty of thus emitting some of its energy. For example, the earth 
loses some of its heat to space continuously day and night, and is 
said to "cool by radiation." Of course, when the sun is shining 
clearly, the illuminated side of the earth is gaining more energy from 
the sun than it is losing to space. The rate of radiation increases as 




the fourth power of the temperature expressed in the absolute scale; 
doubling the absolute temperature of a body results in its sending 
out radiant energy 16 times as fast. 

Characteristics of radiant energy. All electromagnetic waves 
travel through space at the approximate speed of 186,000 miles per 
second. This is generally called the speed of light. The length of a 
wave is the distance between two adjacent crests or pulses. The fre- 
quency is the number of crests that pass a given point per second. 
It follows that the product of the wave length and the wave fre- 
quency is constant, or that the two quantities are inversely propor- 

The properties of the various electromagnetic waves are related to 
their lengths and frequencies. When arranged according to their 
lengths, they form a continuous arrangement known as the electro- 
magnetic spectrum (Fig. 46). At the left end are the extremely short 
waves known as cosmic rays, gamma rays, and x-rays; next in order 
with increasing wave length come the ultraviolet rays, the visible 




A i 








/ / 



















Fig. 46. Diagrammatic Representation of the Electromagnetic Spectrum. The 
figure needs to be greatly elongated to represent the correct ratios of the wave lengths, 
which vary between Kh 10 and 10 10 centimeters. 

light rays, and the infrared or so-called heat rays; and finally come 
the Hertzian electric waves, including those used in radio transmis- 
sion. The visible rays of light have a length extending from about 
3.8 to 7.6 ten-millionths of a meter; the waves used in radio broad- 
casting may vary from less than 10 to more than 30,000 meters. 

The waves received from the sun and those sent out by the earth 
are those with which we are particularly concerned in meteorology. 
The sun's rays include not only the visible light rays, but extend 
from the ultraviolet far into the infrared. Radiation from the earth 


is always in long heat waves. The use of the expression heat tvaves, 
as applied to the long-wave radiation emitted by warm or hot bodies, 
is misleading, since all electromagnetic waves produce heating ef- 
fects when absorbed. 

Transmission, absorption, and reflection. Not only does radiant 
energy travel through space without the presence of any material 
substance, but portions of it also pass through certain kinds of mat- 
ter. Light rays, for example, travel through air, water, and glass, and 
x-rays and other short waves through denser substances opaque to 
visible light. In these cases the radiation is said to be transmitted; it 
is not itself affected and has no effect on the matter through which 
it passes. Most substances show a selective transmission; that is, 
some of the wave lengths get through but others do not. For exam- 
ple, window glass admits the light from the sun, but it does not 
transmit outward the long heat waves originating in the room. Dif- 
ferent substances select different wave lengths for transmission. 

That portion of the radiant energy which enters a substance but 
is not transmitted through it is said to be absorbed. It thereby ceases 
to be radiant energy and is changed into some other form of energy, 
often into heat, but sometimes into the energy used in evaporation or 
in chemical changes. Only the radiation that is absorbed has any ef- 
fect on the object with which it comes in contact. Selective transmis- 
sion implies selective absorption; those wave lengths not selected for 
transmission are absorbed. A radio receiving set illustrates the selec- 
tive transmission and absorption of electromagnetic waves. It has a 
device by which certain wave lengths are selected for amplification 
and others are "tuned out." Some of the waves reaching a material 
surface may be reflected, that is to say, turned back without enter- 
ing the substance. The only result is to change the direction of mo- 
tion of the waves. The reflection may be regular, as from a mirror 
or other smooth surface, or diffuse, like that from the surface of 
the ground. Objects are made visible by reflection; those that reflect 
no light cannot be seen, unless they themselves are emitting light 

Incoming Solar Radiation 

The heat of the atmosphere and of the surface of the earth is de- 
rived almost wholly from the sun. The amounts received from the 
moon, planets, stars, and the earth's interior are negligible in com- 


parison, as is evident when we remember how the temperature 
ordinarily rises by day under the influence of sunshine and falls by 
night, when these other factors are equally as effective as by day. 
That part of the incoming solar radiation that reaches the earth's 
surface is given the special name of insolation. 1 

Solar constant. Although the earth receives but a minute por- 
tion of the total radiation emitted by the sun (approximately 
1/2,000,000,000), owing to the small angle subtended by the earth 
as seen from the sun, the amount received is great. A quantity of 
heat is measured in calories, a calorie being the heat required to 
raise one gram of water from 15C. to 16C. The average intensity 
of the solar radiation is found to be about 1.94 calories per square 
centimeter per minute, at the average distance of the earth from 
the sun, when measured on a surface perpendicular to the solar 
beam, and after making allowance for the loss in passing through 
the atmosphere. This is called the solar constant, so named when 
the intensity was assumed to be invariable within the errors of meas- 
urement. Later investigations indicate that there are slight varia- 
tions in short periods of a few days. There is more conclusive evi- 
dence that there are variations, amounting to about 3 per cent, in 
periods of a few years. The variations are associated with changes 
in sunspot activity. If we assume that in the course of a year, half of 
the energy expressed in the solar constant reaches the earth at lati- 
tude 40, the energy received amounts to more than 5 million kilo- 
watt hours per acre. 

The observations fixing the average value of the solar constant 
and its variations have mostly been made and are being continued 
by the Smithsonian Institution. They are made on mountains in arid 
regions in the southwestern United States and in Chile and Sinai. 
These sites were chosen in order to avoid as much of the dust and 
moisture of the lower atmosphere as possible. Furthermore, meth- 
ods have been devised by which the remaining errors introduced by 
variable losses in the atmosphere can be largely eliminated. These 
observations have been made since 1918 and disclose a probable 
range in the solar constant of not more than 5 per cent. The variation 
is found to be in the blue, violet, and especially the ultraviolet por- 
tion of the spectrum, rather than in the greater wave lengths. 

1 The word insolation is sometimes used more technically to mean the rate at which 
direct solar energy is received at the earth on a horizontal surface. 



Solar radiation measurements. Continuous records of two values 
in connection with solar radiation are now obtained at a number of 
places in various parts of the United States and in other countries. 
One quantity measured is the intensity of direct solar radiation at 
normal incidence, that is, on a surface kept at right angles to the sun's 
rays; the other is the total radiation received on a horizontal sur- 

47* Two f .ifi'a, 

Jttt* tw to an 

to of lil, 


face, including radiation reflected from sky and clouds as well as that 
coming in a direct line from the sun. The instrument used to measure 
solar radiation is called a pyrheliometer, and is based on the thermo- 
electric effect differential heating produces an electromotive force 
that is closely proportional to the amount of radiation received. (Fig. 
47). The resulting current is recorded by a potentiometer. The data 
obtained by the use of these instruments are used by architects and 
illuminating engineers and in various industries. They are also of im- 
portance to botanists and agriculturists, for the growth of plants is 
closely correlated with the amount of solar radiation received. 

Amount of Insolation Received at a Fixed Location 

It is evident that the energy received from the sun at the surface 
of the earth differs from the solar constant and averages consider- 
ably less. The actual amount received at a;iy point depends on the 
following factors: 

Solar constant; atmospheric absorption and reflection. The 
amount of insolation received changes as the solar constant varies, 
that is, as the actual energy emitted by the sun changes, but the per- 
centage change in the solar constant is small at most. Of much 
greater importance in determining the amount received by the earth 
is the fact that this varies as the amount absorbed and reflected by 
the atmosphere varies. Changing conditions of cloudiness, dustiness, 
and humidity of the atmosphere are continually altering the amount 
of radiant energy transmitted to the earth. A large and variable 
amount is reflected back to space by clouds. The most important 
absorbing constituent of the atmosphere is water vapor. 

Distance of earth from sun. Since radiation spreads out spheri- 
cally from its source, the amount intercepted by a given area varies 
inversely as the square of its distance from the source. The distance 
of the earth from the sun averages about 93,000,000 miles ( 150,000,- 
000 km) but varies somewhat during the year because the earth 
moves in a slightly elliptical orbit. The earth is about 3,000,000 miles 
nearer the sun on January 1 than on July L In consequence, the 
total solar energy reaching the entire earth's atmosphere is about 7 
per cent greater in January than in July, although it is evident that 
we of the Northern Hemisphere receive more heat in July than in 
January. The distance factor would lead one to believe that the 


Southern Hemisphere should experience hotter summers and colder 
winters than the Northern Hemisphere. This is not true. Examina- 
tion of a globe will reveal that a much larger portion of the land 
area of the earth is north of the equator. Land heats and cools more 
rapidly than the oceans and thus renders the seasonal variation in 
distance of the earth from the sun of little consequence. Of much 
greater importance in determining the amount of insolation received 
at a fixed location are two other influences next to be considered. 

Length of day and angle of incidence. As the earth moves about 
the sun, its axis maintains a nearly constant direction in space, mak- 
ing an angle of 66 K> with the plane of its orbit. In consequence, 
the angle at which the sun's rays strike a point on the earth changes 
with the changing position of the earth relative to the sun. On June 
21, the sun is vertically overhead at noon at the tropic of Cancer and 
has its greatest noon elevation at all latitudes north of the tropic 
and its least elevation at all points in the Southern Hemisphere. Six 
months later, on December 21, the relative positions of the hemi- 
spheres are reversed and the noon sun is directly overhead at the 
tropic of Capricorn (Fig. 48). These dates are the summer solstice 
and the winter sohtice, respectively, in the Northern Hemisphere. 
The time of winter and summer are reversed for the Southern Hemi- 

As time progresses from December 21, the vertical rays of the 
noonday sun move northward, and are directly overhead at the equa- 
tor on March 21. They continue northward to the tropic of Cancer, 
arriving there on June 21, and return to the equator about Septem- 
ber 22. These dates are called the vernal equinox, summer solstice, 
and autumnal equinox, respectively. At equinox, days and nights are 
of equal length throughout the world. At other times of the year, 
days and nights are not of equal length except on the equator. The 
dates of equinoxes and solstices vary at times by one day or so be- 
cause our calendar is not an exact representation of the length of a 
solar year. 

The variations in the angular elevation of the sun produce changes 
in the amount of insolation received. A given surface receives most 
rays when they fall perpendicularly upon it. The same amount of 
incoming radiation is represented by the two lines, A, in Fig. 49. 
When the rays are perpendicular, they cover an area one of whose 
sides is A, but when the elevation of the sun is represented by the 




JUNE 21 


DEC. 21 

SEPT. 23 

Fig. 48. Earth's Orbit About the Sun. 

angle x, they cover the greater area whose side is A + B. The same 
amount of insolation is thus spread over a greater surface, and the 
energy received per unit area is less in the ratio of A to A + B. The 
angle of incidence is defined as the angle which the sun's rays make 
with the perpendicular. In the case of vertical rays, the angle Is zero; 
in the case of the slanting rays in Fig. 49, the angle of incidence is 
90 x. The angle of incidence at solar noon for any place on the 
earth can be determined by the following rule: The number of de- 
grees in the angle of incidence is equal to the number of degrees of 
latitude between the observer and the latitude tvhere the sun's rays 
are vertical. As the angle of incidence increases, the amount of in- 
solation decreases if other variables do not interfere* 


UMIT or AI* 


Fig. 49. Effect of Angle of Incidence on Insolation. Disregarding atmospheric ab- 
sorption, the intensity of insolation when the sun has an angular elevation of x, com- 
pared with the intensity when the sun is vertical, is the ratio A to A 4- B, which is *fri 
x. The length of path when the sun is vertical, compared with the length when the 
sun's elevation is *, is tho ratio C to D, which is also sin x. 

There is also a secondary effect of the angle of incidence on the 
amount of radiation received at the earth's surface. As the inclina- 
tion of the rays from the vertical increases, the length of their path 
through the air increases, in the ratio of C to D, which is the same 
ratio as A to A -f B. The longer the path through the air, the greater 
is the absorption and scattering by the air, especially the lower air. 
Hence, when the sun is near the horizon, its effect is weakened not 
only by the spreading out of the rays but also by the loss of heat in 
passing through much moist and dusty air. Whether the losses due 
to absorption are more important than those due to inclination de- 
pends upon the condition of the atmosphere. It has been found that 
at Montpelier, France, 71 per cent of the incoming solar radiation 
reaches the earth in December, and only 48 per cent in the summer 
months. There the increased amount of moisture (specific humidity) 
in the summer air reduces the seasonal temperature several degrees. 

Not only is the angular elevation of the sun greater in summer 
than in winter, but the duration of sunshine is greater, the days are 
longer, and the nights shorter. It is evident that, other factors being 
equal, the amount of insolation received is directly proportional to 
the length of time during which it is being received. This fact has 
a very important effect in increasing summer insolation in the mid- 
dle and higher latitudes. The effect increases toward the poles, where 
the sun is continually above the horizon in the summer, and de- 
creases toward the equator, where the days and nights are always 
of the same length. See Table III. At latitude 40 there are about 15 


hours of possible sunshine in midsummer as compared with about 9 
in midwinter. On June 21, the North Pole receives more energy from 
the sun than does a point on the equator on the same date, but the 
total amount received during a year at either pole is only about 41 
per cent of that at an equatorial location. This statement considers 
only the effect of position on the earth with respect to the axis of 
rotation. Actually, lines of equal insolation do not follow the circles 
of latitude exactly because of many local influences, such as prevail- 
ing cloudiness, to be discussed later. An important practical result, 
however, of the long summer days in high latitudes is that wheat 
and other crops can be grown far poleward, in spite of a very short 
summer, because of the great amount of sunshine during the brief 
growing season. 


H>ad down for the Northern Hemisphere. 


March 21 

June 21 

Sept. 23 

Dec. 21 


12 hrs. 

12 hrs. Oinin. 

12 hrs. 

12 hrs, min. 


12 hrs. 

12 hrs. 35min. 

12 hrs. 

1 1 hrs. 25 min. 


12 hrs. 

13 hrs. 12min. 

12 hrs, 

10 hrs. 48 min. 


12 hrs. 

13 hrs. 56min. 

12 hrs, 

10 hrs. 4 min. 


12 hrs. 

14 hrs. 52 min. 

12 hrs. 

9 hrs. 8 min. 


12 hrs. 

16 hrs. 18 min. 

12 hrs. 

7 hrs. 42 min. 


12 hrs. 

18 hrs. 27 min. 

12 hrs. 

5 hr*. 33 min. 


12 hrs. 

2 months 

1 2 hrs. 

Ohm. Omin. 


12 hrs. 

4 months 

12 hrs. 

hrs. min. 


12 hrs. 

6 months 

12 hr*. 

hrs, min. 


Sept. 23 

Dec. 21 

March 21 

June 21 

Read up for Southern Hemisphere 

Direct Effects of Solar Radiation 

The energy coming to the earth from the sun in the form of solar 
radiation is either absorbed or reflected or scattered. Varying con* 
ditions over the face of the earth and in the atmosphere cause the 
distribution of energy received to vary with time and with geo- 
graphical location. 


The ratio at which light rays are reflected from a surface in com- 
parison to the total rays striking the surface is called the albedo of 
that body. The reflectivity of the earth's surface, assuming average 
cloud cover and other atmospheric conditions, increases from the 
equator to the poles and averages approximately 34 per cent. 1 The 
earth is then said to have an albedo of 0.34. Clouds have greater 
albedo than land surfaces, averaging about 0.55 and 0.10, respec- 
tively. In addition to those solar rays which are reflected, some are 
scattered by dust particles and other impurities in the atmosphere, 
some are absorbed by the atmosphere, and the remainder are ab- 
sorbed by the earth's surface. About 19 per cent of solar radiation is 
absorbed in the atmosphere and about 47 per cent is absorbed at 
the earth's surface (Fig. 50). Assuming that the earth and the sur- 
rounding atmosphere become neither warmer nor colder over a long 
period of time, all of the heat absorbed must eventually be reradi- 
ated back to space. 

Effect on air. Most of the reflection from the atmosphere is from 
the upper surface of clouds. Clouds have a high reflective power 
which varies with their thickness and the amount of liquid water 
they contain. Clouds of ordinary thickness and density probably re- 
flect from 75 to 80 per cent of the incident radiation. The greater 
part of this reflection is directed outward and so is lost to us, but a 
part of the reflection is from the upper surface of one cloud to the 
lower surface of a higher cloud, and thence to the earth. Some of the 
heat and light we receive on partly cloudy days consists of this re- 
flected radiation. The earth and lower air eool less rapidly on a 
cloudy night than on a clear night, not because radiation from the 
earth is less rapid, but because much radiation is reradiated from 
the clouds to the earth, If the air were free of dust and moisture, 
only a small fraction of solar radiation would l>e absorbed; nearly all 
would be transmitted to the earth without alteration, Such air would 
hardly be perceptibly heated by sunshine, for it is only absorbed 
radiation that increases temperature. 

Oxygen and nitrogen are practically transparent to the sun's radia- 
tion, and such absorption as occurs in air, except by dust and mois- 
ture, is accounted for largely by carbon dioxide in the long-wave 
"heat rays** of the infrared, and by ozone in both the ultraviolet and 

1 Henry C. H might on, "On the Annual Heat Balance of the Northern Hemisphere," 
Journal of Meteorology, Vol. 11 ( 1954), pp. 1*9. 


29% / 



^ % 
















Fig. 50. Diagrammatic Representation of the Distribution of Solar Radiation. (After 
//. G. H outfit on. ) Of the incoming radiation, some? is absorbed by the atmosphere, 
some is reflected back to space, and the remainder is absorbed by the earth. The heat 
balance is maintained by the earth loosing its heat back to space through lonjj-wave 
radiation, latent heat, and sensible heat. 

the infrared waves. Much the greater part of the absorption by the 
gases of the air is by water vapor, which absorbs most of the long- 
wave radiation. Since the radiation emitted by the earth is all long- 
wave, the water vapor absorbs a much greater proportion of earth 
radiation than of solar radiation, and thus acts as a trap to conserve 
the energy received from the sun. Coincident with these absorption 
processes, reradiation from the atmosphere is active and continuous. 
Solid particles of dust and smoke in the air, and liquid or solid 
particles of water in the air, absorb and reflect considerable, but ex- 
tremely variable, amounts of solar radiation. The dry air of deserts, 
if it is not filled with dust, absorbs little radiation, and hence the 
sun has an intense heating effect on solid objects. Elevated regions 
are above much of the dust and moisture of the air, and conse- 
quently there is little absorption by the air above them. Hence the 
air remains cold, and the sun's rays have much energy left to be ab- 
sorbed by objects at the surface. That is why one is often warm on 
winter days in the mountain sunshine, but cold in the shade, the 
difference between sunshine and shade being much greater than in 


Effect on land surfaces. Of the radiation that gets through the 
atmosphere and reaches the surface of the earth, a part is reflected 
back into the atmosphere and the remainder is absorbed at the sur- 
face. The proportion that is reflected by land surfaces varies greatly 
with the condition and color of the surface. If the land is covered 
with grass or trees, or is black, cultivated soil, the reflection may be 
in the neighborhood of 10 per cent. A bare, hard, sandy soil may 
reflect 20 per cent, and freshly fallen snow 70 to 80 per cent of the 
incident radiation. 

The remaining percentage of the radiation that reaches the earth 
is there absorbed and changed into other forms of energy. It is thus 
that the soil is warmed. Land surfaces are good absorbers and there- 
fore heat rapidly when the sun is shining on them. Moreover, heat 
does not penetrate deeply into the soil but remains in a thin surface 
layer, For this reason also the surface heats rapidly. The daily varia- 
tion is small below 4 inches, but a slight daily change may occur to 
a depth of about 3 feet. 

A good absorber is a good radiator, and the land surface that 
warms rapidly by day eools rapidly by night. Thus a large part of 
the heat received is soon returned to space. A sandy desert soil exem- 
plifies the maximum of rapid changes in a thin layer. Under such 
conditions, a change of 49 F. from clay to night has been observed 
in the surface layer of soil, whereas the change was only IF. at a 
depth of 16 inches. Moist soil does not heat so rapidly as dry soil 
because some of the energy received is used in evaporating the 
water* and because it takes more energy to heat water than soil. 

Snow reflects much solar radiation. It absorbs the remainder and 
also absorbs or reflects downward the radiation from the soil. A good 
snow cover protects the land from large daily changes and often 
prevents its freezing through severe weather. In the spring when the 
snow disappears, after having persisted for a considerable period, 
it is often found that the ground beneath is unfrozen, even where 
near-by bare ground is still frozen hard. In such cases the snow acts 
as a blanket; heat is conducted from lower ground levels to the 
surface, but it cannot escape to the air and therefore keeps the sur- 
face layer of soil warm. 

With the increased insolation of summer, outside the tropics, not 
all the heat stored by day is lost in the short nights, and the soil 


becomes progressively warmer. In winter more heat may be lost by 
radiation through the long nights than is received by day. A land 
surface therefore becomes hot by day and in summer, and cold by 
night and in winter. It does not store a large amount of heat. 

Effect on water surfaces. Reflection of insolation from a water 
surface depends upon the smoothness or roughness of the water, and 
especially upon the angle of incidence of the sun's rays. When the 
sun is not more than 5 above the horizon, 40 per cent of the insola- 
tion may be reflected. As the sun's elevation increases, the percen- 
tage reflected steadily decreases until not more than 3 per cent or 4 
per cent is reflected when the sun is 50 or more above the horizon. 
On the average it is probable that the reflection from water surfaces 
is about the same as from land. Hence, the absorption is about the 
same. But water surfaces and land surfaces respond quite differently 
to the absorbed insolation. 

There are four important ways in which the effects differ: (1) 
Radiation penetrates to much greater depttis in water than in land, 
More than one-third of it reaches a depth of 3 feet, and about one- 
tenth is transmitted to 30 feet. Small amounts of light have been 
observed at depths of 1,700 to 1,900 feet (520-580 in) beneath the 
surface of the sea. ( 2 ) Of even greater importance in distributing the 
heat through a considerable depth are the wave movements and 
general turbulence of the sea surface. Because of this mixing of the 
waters, and because 90 per cent of the absorption is in the first 30 
feet, it may be assumed that the heating effect is uniformly dis- 
tributed through a 30-foot layer, Thus, there is* a great volume of 
water to be heated as compared with a 4-inch layer of land. In ad- 
dition, the heat absorbed is often transported great distances by cur- 
rents, tides, and other movements of the water. The cooling of the 
surface water by night also contributes to the mixing effect. As the 
water becomes cooler, it becomes denser; it sinks and is replaced 
by wanner water from below. 

The two other reasons why insolation has comparatively little ef- 
fect on the temperature of water surfaces are: (3) A large part of 
the energy absorbed by the water, probably about 30 per cent of it, 
is used in evaporating the water and is therefore not available for 
raising its temperature. Furthermore, evaporation increases the salin- 
ity, and hence the density, of sea water and thereby contributes to 


the mixing. Evaporation is greatest when the water is warmer than 
the overlying air, l>eeause then the vapor pressure at the surface 
is greater than in the air. (4) The specific heat of water is greater 
than that of other natural substances. To raise the temperature 
of one pound of water IF. requires three times as much heat energy 
as to heat one pound of soil IF. 

For all these reasons, water areas heat slowly, store much energy, 
and cool slowly. They are great storehouses of heat. Large land 
ureas have great and rapid temperature changes and little storage 
capacity. The oceans are conservative; the continents, radical. This 
difference is of fundamental importance in meteorology and climatol- 
ogy, as we shall find. It should he noted, however, that an ice- 
covered body of water acts much us a snow-covered land surface. 
It reflects a high percentage of the incident radiation; it warms lit- 
tle by clay and cools rapidly by radiation at night. 


One primary cause of temperature changes in the lower air is con- 
duction of heat to or from the earth's surface. Conduction is the 
process by which heat is transferred through matter, without trans- 
fer of the matter itself. When one end of a silver spoon is heated, the 
other end soon Ixxxmies hot by conduction; but when one end of a 
piece of wood is heated, the other end remains cool. Silver is a good 
conductor of heat; wood, a poor conductor. Conduction is always 
from the warmer to the colder point. On a sunny day, the earth's 
surface is warmed by absorbing insolation, and then, after the earth's 
temperature has increased above that of the air, the air in contact 
with it is wanned by conduction as well as by radiation. Similarly, 
at night, the first process is the cooling of the ground, and then the 
cooling of the air as it conducts and radiates some of its heat to the 
ground. Thus, air tends to have the same temperature as the sur- 
face with which it is in contact. Air is a poor conductor, however, 
and the actual conduction during the course of a day or night af- 
fects only two or three feet of air. Wind and turbulence, however, 
bring fresh air in contact with the surfaces and distribute the warmed 
or cooled air to a considerable height. The exchange of heat by con- 
duction is less than that by radiation* 


Air temperatures. The temperature of the air lags behind that of 
the earth and changes less; the air is not so warm as the land, in the 
sunshine, nor so cool under outgoing radiation conditions at night, 
This fact applies to air locally heated or cooled not to warm or cold 
air that may be brought in from other regions. The poor conductivity 
of the air and its slow Ibss of heat by radiation explain why frosts 
sometimes occur when the general air temperature is considerably 
above freezing. The grass, the paving, and other surfaces where frost 
forms are colder than the mass of air a few feet above them. It is 
clear also why a thermometer must l>e sheltered from radiation, 
direct or reflected, if it is to assume by conduction the temperature 
of the surrounding air. 

The question, "What is the temperature in the sun? M meaning in 
sunshine, has no answer. Each different object exposed to the sun's 
rays absorbs radiation differently and takes on a different tempera- 
ture. Black objects become warmer than ligltf-colorcd ones, and dry 
ones warmer than moist ones. When a black-bulb thermometer, that 
is, a thermometer with a large bulb coated with lampblack, is ex- 
posed to solar radiation, it often has a temperature 60F. or 70F. 
higher than that of the surrounding air. A piece of black fur exposed 
to the sun's rays in winter in the Alps reached a temperature of 140 
when the air temperature was 41'. On the other hand, such an ob- 
ject gets much colder than the air at night. It is the temperature of 
the air, and not of absorbing and radiating solid bodies, that is of 
primary concern in discussions of the weather. 

Earth temperatures. As previously stated, the heating of a land 
surface by insolation is confined to a thin surface layer. This is so 
because land is a poor conductor. Heat is conducted downward so 
slowly that the diurnal change of temperature ordinarily penetrates 
but two or three feet into the soil Before it has reached that depth, 
night has arrived, and the surface is cooling. The annual variation 
disappears in all latitudes at a depth of only a few feet. The ampli- 
tude of the change in this surface layer diminishes rapidly with depth 
and depends largely upon the seasonal differences in air tempera* 
ture. The temperatures at 100 feet below the surface vary with the 
latitude, being dependent on the mean annual temperatures of the 
air at the surface. At greater depths, the temperature of the earth 
increases slowly but not uniformly. 



Conduction and the absorption and emission of radiation are proc- 
esses which originate temperature changes in a substance, whether 
solid, liquid, or gaseous. Another method of transferring heat is of 
great importance in relation to the temperature and behavior of the 
air, although it is not an original source of gain or loss of heat energy. 
This is convection, the transfer of heat by internal mass movements 
of the substance containing the heat (Fig. 51). Such movements 
result from temperature differences within the substance, and can 
occur only in liquids and gases, not in solids. As used in meteorol- 

Fig, 51, Transfer of Heat by Convection. A, water heated at the top while lower 
portion remains cool. B, water heated at the bottom; all becomes nearly equally 

ogy, convection refers to vertical movements of the air. Advection 
is the term used with reference to the horizontal transport of heat 
by winds and slow drifting movements of the air. 

Convection in a liquid. If a test tul>e is filled with water, and a 
flame is applied near the top, the water at the top may be brought 
to the boiling point while that at the bottom is relatively cool. In 
this case, the lower water is heated only by conduction. If the flame 
is applied near the bottom of the tube, the heated water expands 
and is displaced by the cooler, denser liquid above it, setting up a 
convectional circulation, in which heat is transferred by the move- 
ment of the water. Thus, the entire mass becomes heated to the 
tx>iling point at nearly the same time. Fig. 52 illustrates a convec- 
tional circulation set up in a vessel of water heated at the bottom 
over a small part of its area. 


Convection in the air. Gases 
move even more freely than do 
liquids and likewise expand when 
heated, thereby becoming lighter 
than before, volume for volume. 
Familiar examples of convection in 
the air on a small scale are the draft 
up a chimney and the rising of air 
over a heated radiator. In these 
cases air is warmed at the bottom, 
and colder, heavier air pushes it Flfr 52 Convoctiona1 circulation in 
upward out of the lowest place. * Liquid. 

The effectiveness of a warm-air furnace in heating a house depends 
upon this method of transferring heat. We have noted that, while 
the earth's surface is heated by absorption of insolation, it is for 
many reasons very unequally heated, and that the lower air is heated 
by radiation and conduction and likewise unequally heated. We 
should therefore expect to find convoctional currents, involving 
downward and upward movements of the air, between areas of con- 
trasting temperatures, as, for example, between the oceans and the 
continents, and on a large scale between equatorial and polar re- 
gions. It will be seen, as we proceed, that such movements are of 
primary importance in the study of the atmosphere. 


1. Tie a string around a globe to represent the* boundary between the 
light hemisphere and the dark hemisphere on June 21. Measure several 
latitude circles, including the equator, to determine the relative lengths 
of day and night on that date. 

2. Repeat the alxne exercise for (a) September 23, (b) December 21, 
and (c) March 21. 

3. Why are days and nights always of equal length at the equator? 

4. What is the maximum difference in the noon elevation of the sun 
(expressed in degrees) during a year at any point outside the tropics? 

5. Explain how a navigator can use the sun to determine hte latitude 
if he knows the date and the angle of incidence of the noon sun. 

6. Find the exact latitude of your school, and on the day this problem 
is assigned, measure the angle of incidence of the sun at noon with a pro* 
tractor. At what latitude are the sun's rays vertically overhead? 



We have seen that the sun is practically the only source of heat 
for the atmosphere. It is now necessary to consider how the atmos- 
phere reacts to heat energy. All air does not have the same prop- 
erties, nor does the same air necessarily maintain its properties for 
any given period of time. The atmosphere may be compared to a 
very temperamental person whose response to a stimulus is quite 
unpredictable unless one is thoroughly acquainted with the charac- 
teristics of that individual. 

You have learned many of the atmosphere's characteristics. The 
gas laws of Boyle, Charles, and Gay-Lussac were discussed in Chap- 
ter 1. Combined, these laws express a universal characteristic of the 
atmosphere: Pressure times the volume equals a constant times the 
temperature (PV = RT). In the last chapter, it was shown how 
heat is transferred to and by the atmosphere through radiation, con- 
duction, and convection. Convection unlocks other potential forces 
of the atmosphere, and we shall proceed to study these in more 

Adiabatic Temperature Changes 

When air ascends, the pressure on it decreases, and the gases ex- 
pand according to the gas laws. Expansion constitutes work in the 
physical sense and uses energy. The energy expended is heat energy, 
and the effect is to cool the air. Almost everyone has observed this 
temperature phenomenon at one time or another. The tube of a 
bicycle or automobile tire-pump will become heated as air is forced 
by the piston into the tire, or frost may occur on a warm summer 



day about the escape-valve of a pressure tank when compressed air 
is being released. 

Ascending air cools as it expands under decreasing pressure and 
descending air is wanned by compression as it comes into regions 
of greater pressure. Note that these tempertaure changes are not 
related to any transfer of heat to or from the air. The air becomes 
cooler or warmer without any conduction or radiation. The change 
in temperature is an internal change as a result of the change of 
pressure upon it. Such changes in temperature are called adiabatic 
changes, the word implying "without transfer of heat." It may be 
shown by a mathematical discussion of the properties of gases that 
when dry air rises above the ground surface, the dynamic cooling 
due to expansion is at the nearly uniform rate of 5.5F. per 1,000 
feet, or 1C. per 100 meters. The rate of warming with descent is 
the same. This is the adiabatic rate for unsaturated air. 

When considerable moisture is present in rising air, the cooling 
caused by rising and expansion may result in saturation and then 
in condensation of some of the water vapor. The level at which con- 
densation begins is known as the lifting condensation level ( LCL ) . 
Beyond this point, two factors influence the air temperature. The 
dynamic factor continues to cool the air adiabatically, but the re- 
lease of latent heat of condensation tends to warm the air ( see pages 
42-43). The net result is a retarded rate of cooling. Thereafter, while 
condensation continues, rising air cools at a retarded adiabatic rate. 
This retarded rate of cooling is called the saturation adiabatic rate 
or wet adiabatic rate. It is not so nearly constant as the dry adiabatic 
rate, but depends on the temperature and pressure, and on how 
much of the condensed moisture is carried along with the rising air 
or precipitated out of it. Under changing conditions, the wet adi- 
abatic rate varies from about 0.4C. to nearly 1C. per 100 meters. 
The average value for warm temperatures is about 0.5C. per 100 
meters or 3F. per 1,000 feet. 

Descending air is dynamically warmed by compression. The pos- 
sible moisture content of the air is thereby increased. It follows that 
there will be no more condensation, but, on the contrary, there will 
be evaporation if liquid water is present in such air. This evapora- 
tion will retard the warming of the descending air. If evaporation 
is sufficient to maintain saturation as the air descends, its rate of 
warming is a wet adiabatic rate. If there is no condensed moisture 



present, descending air warms at the dry adiabatic rate. Usually, 
when condensation has occurred, some condensed moisture is pres- 
ent when the air begins its descent, but this disappears as the tem- 
perature rises, and for the remainder of its descent the air is un- 
saturated (Fig. 53). If condensation has occurred during the adia- 
batic processes, the air returns to its point of origin somewhat warmer 
than the original temperature. 

10,000 FEET 

5,000 FEET 




















S9* 40* 4ft* 60* 69* CO* fS* 70* 79* SO* 89* tO* 99* 100 


Fig. 53. Adiabatic Changes of Temperature. If a parcel of air at sea level, A, with 
a temperature of 80 F., be lifted 5,000 feet before becoming saturated, it would 
have a temperature at that level of 52 l /zF. On being lifted to 10,000 feet, the same 
parcel would have a temperature of approximately 37 1 /2F. On returning to sea level, 
D, this parcel of air would have a temperature of 92V2F. Drawing by E. L. Peterman. 

Note that adiabatic changes occur only when air is expanding or 
being compressed, and that they occur without any heat being 
added to or taken from the air. Let us consider whether such a con- 
dition ever occurs in nature. Take, for example, a large mass of ris- 
ing air, such as that which precedes the formation of a cumulus 
cloud. A small part near the outer surface of the rising mass may 
have its temperature affected by mixing with the surrounding air; 
also, near the outer surface there is some interchange of heat by 
radiation and absorption. But air is a poor conductor and a poor 
absorber, and these modifications do not reach any great distance 
into the interior of the rising column. Much the greater portion of 
the ascending air is subject to no appreciable loss of heat to, or gain 
from, the outside. Hence, its changes in temperature are essentially 


adiabatic, the result of expansion under decreased pressure, and are 
treated as such for practical purposes in interpreting the behavior 
of the cloud. 

Potential temperature. We have seen that the actual tempera- 
ture of a sample of air varies greatly as it responds to changes of 
pressure. Sometimes it is desirable to recognize and make use of a 
more conservative temperature value, 

Suppose that a quantity of unsaturated air at or near the surface 
of the earth and subject to a pressure of 1,000 millibars has a tem- 
perature of 70F. If it is now forced to rise 1,000 feet and remains 
unsaturated, it cools at the dry adiabatic rate and its temperature 
will be reduced to 64.5F. (70 -5.5). If it then descends to its 
original level where the pressure is 1,000 millibars and warms adia- 
batically, it will return to its original temperature of 70F. No mat- 
ter how far this quantity of dry air rises or descends, or what pres- 
sure changes it undergoes, it will always return to a temperature of 
70 when it returns to a pressure of 1,000 millibars, provided that 
it remains unsaturated and is subject only to dynamic influences. In 
general terms, potential temperature is defined as the temperature 
that a quantity of air would have if brought by dry adiabatic 
changes to a pressure of 1,000 millibars. Potential temperature is the 
actual temperature reduced adiabatically to a standard pressure. 
Changes of elevation result in changes in the existing temperature 
but make no difference in the potential temperature of unsaturated 

Suppose, now, that the air, having risen 1,000 feet and reached a 
temperature of 64.5F., becomes saturated at that point, but con- 
tinues to rise another 1,000 feet, with condensation and precipita- 
tion occurring. During the second thousand feet of rise, it will have 
cooled only 3, and at the top will have reached a temperature of 
61.5F. If it now descends to its original level, warming all the way 
at the dry adiabatic rate, when it reaches the standard pressure its 
temperature will be 61.5 + 11.0 = 72.5F. instead of the 70 at 
which it started. We see that the potential temperature may be in- 
creased by adiabatic processes, when these involve condensation. 

Equivalent potential temperature. An even more conservative 
property of a parcel of air is its equivalent potential temperature. 
Again let a quantity of air, starting at the standard pressure of 1,000 
millibars, rise until it has lost all its moisture, cooling first at the dry 



adiabatic and later at a retarded rate, and absorbing the heat re- 
leased by condensation. This is a pseudo-adiabatic process equiva- 
lent to adding to the air the heat of condensation latent in all its 
vapor. Let this completely dry air then descend to the standard 
pressure. The temperature which it then assumes is called its equiva- 
lent potential temperature and can be calculated when the original 
temperature and humidity are known. 

The potential temperature of a given mass of air is unchanged by 
dry adiabatic processes. The equivalent potential temperature re- 
mains the same even though the vertical movements involve con- 
densation and precipitation. It can be changed only by the addition 
of water vapor through evaporation, or by the loss or gain of heat to 
or from outside sources. 

In Fig. 54, if air at 1,000 millibars pressure and 18C. (64.4F.) 
rises and cools at the dry adiabatic rate to the 857-mb level (4,500 
feet), where it becomes saturated, its potential temperature is still 
18C. If it then cools at the retarded rate to the 700-mb level 
(10,000 feet), its potential temperature becomes 25C. If it con- 
tinues to rise until all its moisture is condensed and the latent heat 
absorbed, and then returns adiabatically to 1,000 mb pressure, its 
temperature will be 36C. This is the equivalent potential tempera- 
ture of the air. 

The adiabatic chart. One of the very important tools of the me- 











IS* -10* -5* 0" 5' 10* 15" 20* 25" 30* 35" 40* 


Fig. 54. The Mechanics of Equivalent Potential Temperature. 



teorologist is the adiabatic chart. It is simply a graphical means of 
solving many complicated mathematical relationships which exist 
among the properties of the atmosphere. There are many variations 
in the manner in which a chart may be constructed, but always the 
ordinate is some function of pressure ( decreasing upward ) and the 
abscissa is temperature increasing toward the right (Fig. 55). The 
primary horizontal lines then become lines of equal pressure and 
the vertical lines are lines of equal temperature. 

-40 -30 

-20 -10 



20 30 















r\ \ \ 




\ \ 

xjTfc \ ' ; . 



\ \ 

4^j^/ \ >^ , 




M X^ L, 


. \\ K\ ^ 


) \ . 



\ ' \ ; \x '. 

\> \ 


\ * 

\ )C\N 

\ V 


^ \ ' \ N \ 

, \ 

\ '* 


' ^ 

\\ \\ \ 


\ N 





i Vx " 

^ ^N 

s '. 


\ \ 



\ ^ \i 


\ \ 

\ v 


\ \ 

\ \ 




jt___,-_ ^ ^ 

V \\ N\ 


*i ' \ 

\ \ 



1 \ 

- - 


\\ \ N \ \ i ' v 

^ \ 


', ; 

3 x \ \ <N> \ v 

"s V N 

* \ 


\ \ 

\ % Nj \\\ 

\ \ > 


I M^ 


\ \ 

\ \ 

v XN ' \ 

\ K 

, x 

\ 'K 

\ \ 

X V' -\ 


x \\ 

\ \ \ \\ N 
\\ \\\ 

\ \ \ 

\ *<p 1\ l y 

,x. \ - 

\ > 


v v% 


, \\\\\ 

^. N '. 

\ \ 


\ t 

v\ \ 


\ \ \ \ 

\ V 

V ; 

\ \ 

\ 1 

\ \\ 


\'S \\ 

\ \\ 


v \ 

1, \ 

v\ >> 

\ \ 

\ i^N 

\ \ 1 

\ \ \ 

\ y 

^\ H\ x 1 

N\_ N H'V N 

\ \\ 

\ \ i 


\t" ~~^ 


\ V 

\ t V 

\ \N 


\ \ 

\ v \ r\~" \ 1 \ T" 

* V- I ^^ \ >V \ 

i N \ ! l \\ I N 

\ \\ \ N ^' *. x \ 

\ \ \ i 


\ N 


\ \ 
\ I 


.\\ >v '. V ' 

^ R\\ I\M 



-40 -30 -20 -10 10 20 30 40 


Fig. 55. The Adiabatic Chart. Drawing by Herbert Beaufon. 


Three additional sets of lines complete the chart. They are drawn 
at chosen intervals for convenience, but not so numerous as to im- 
pair legibility. Each set of lines represents a function which is de- 
pendent on temperature and pressure. If a given function is to be 
traced from a point not lying on a printed iso-line of that function, 
the constant value is simply parallel to the function line. 

1. Dry adiabats, or lines of equal potential temperature. They 
are the straight dashed lines sloping from upper left to lower right 
and graphing the dry adiabatic lapse rate of 5.5F. per 1,000 feet 
(I.C. per 100m). 

2. Saturation adiabats, or lines of equal equivalent potential tem- 
perature. They are the curving dash-dot lines varying slightly to the 
right of the dry adiabats and numbered in A. according to the 
equivalent potential temperature of a saturated parcel of air situ- 
ated on the line. 

3. Lines of constant saturation mixing ratio, or lines of equal spe- 
cific humidity, in parts per thousand, of saturated air. The almost 
vertical orientation of these dotted lines shows that temperature is 
a more important factor than pressure in determining the capacity 
of the atmosphere to hold water vapor. 

Sometimes lines of equal height are also drawn because they do 
not coincide exactly with lines of equal pressure. For practical ap- 
proximations, however, the 1,000-mb level is at or near the surface, 
850 mb is about 5,000-feet elevation ( 1.5 km), 700 mb is near 10,000 
feet ( 3 km ) , and the 500-mb level is almost 20,000 feet ( 6 km ) above 
sea level. Soundings from radiosondes and other sources are plotted 
on the adiabatic chart to aid in the analysis of the characteristics 
and potentialities of the upper atmosphere. 

Lapse Rates 

We have seen that ascending or descending air changes tempera- 
ture at a definite rate as a result of the changing pressure upon it. 
This does not mean that the overlying air always grows colder at 
these rates. There are many reasons why the rate of change of tem- 
perature of the air above a given area at a given time should only 
rarely coincide with the adiabatic rate of change. In the first place, 
air is not always rising or falling, and therefore not always changing 
adiabatically. Second, air is constantly gaining and losing heat by 


radiation, absorption, and conduction, and often also by evaporation 
and condensation. Third, horizontal movements bring warm or cold 
air from other sources. For these reasons, the real vertical distribu- 
tion of temperature is frequently quite different from that caused 
by adiabatic processes. 

The actual change of temperature with elevation, whatever it may 
be, is called the lapse rate of the air. Vertical temperature gradient 
expresses the same idea, but "lapse rate" is a shorter and more con- 
venient term. The word lapse means in this connection the gradual 
passing from a higher to a lower temperature. In case the air grows 
warmer with increasing height, the lapse rate is negative. Lapse 
rate is the general term; adiabatic and saturation adiabatic changes 
are particular lapse rates occurring under special conditions. 

Variability of lapse rates. During the past fifty years, a great 
many records of the temperature of the air within a few miles of the 
surface of the earth have been obtained by means of balloons, kites, 
airplanes, and, more recently, radiosondes. The lapse rates found in 
different individual ascents have great variability from day to day 
and at different levels in the atmosphere on the same day, especially 
in the first two miles ( 3 km ) . Beyond two or three miles ( 3-5 km ) , 
the rates are likely to be more nearly uniform. That is to say, the 
temperature of the air below two or three miles changes very irregu- 
larly, being influenced by irregular wind movements from various 
sources. These movements sometimes cause a temporary lapse rate 
that is greater than the dry adiabatic rate, and, on the other hand, 
the lapse rate is often less than the wet adiabatic ( Fig. 56 ) . 

In some cases, instead of decreasing with altitude, the tempera- 
ture increases, as shown in the curves just mentioned. Such a con- 
dition is called an inversion of temperature, or simply an inversion. 
On calm, clear nights, as the soil cools rapidly by radiation, the air 
near the surface is cooled both by radiation and by contact with 
the cold earth. Thus it often becomes colder than the air higher up. 
Inversions, or negative lapse rates, frequently occur in this way, but 
they also occur at higher levels, by reason of winds from different 
directions of differing temperatures. 

Observations of air temperatures aloft have now been made in 
many parts of the world and are sufficiently numerous to establish 
a fairly definite normal or average value. This average lapse rate 
is found to be about 3.3F. per 1,000 feet (0.6C. per 100 in) in the 








LAT. 25 N.- LONG. 26 W. 









\ % 



^ v - 



l\) l\> 
to o> 4^ OD f\> 0> C 

















v x 

s \ 

x \ 



\ s 


N ^ 

v ^\ 




X N 

M^, S 


} X 


A v 


s N 


\ \ 






\ N 

> XN 

\ ^ 





N \\ 




x x N 




60 -50 -40 -30 -20 -10 10 20 


Fig. 56. Examples of Lapse Rates. 


lower levels of the atmosphere. It increases slightly with higher ele- 
vations until the outer limit of the troposphere is reached. It is easy 
to confuse the ordinary decrease in temperature with altitude and 
the adiabatic temperature changes resulting from vertical air move- 
ments. The lapse rate of the air is ordinarily less than the dry adia- 
batic rate and about equal to the saturation adiabatic rate. Nor- 


mally the lapse rate is less when the pressure is high than when it is 
low, and less in winter than in summer. 

Stability and Instability 

Stability. The word stability is used in studying the weather to 
indicate a condition of equilibrium. For example, a ruler lying flat 
on the table is in stable equilibrium; if one end is raised and then 
released, it returns to its original position. A ruler standing on one 
end is in unstable equilibrium; if the upper end is moved slightly, 
the ruler does not return to its former position, but takes a more 
stable position. Let us examine the effect of various lapse rates upon 
the stability of the air, that is, upon its tendency to move up or down 
or remain in the original position. 

It is evident that when a certain mass of air is heavier than the 
surrounding air, it will tend to fall or settle downward; if lighter 
than the surrounding air, it will be displaced upward by the heavier 
air. If the temperature of the air is exactly the same throughout the 
first 1,000 feet above the ground, for instance, the air at the bottom 
is a little denser because of the added pressure upon it, and hence 
a little heavier, volume for volume, than that at the top. Because it is 
heavier, it tends to stay at the bottom; but suppose by some means 
we force a certain portion of it to rise through the surrounding air. 
This rising portion cools at the adiabatic rate. Therefore, at any level 
to which the rising portion ascends within this assumed layer of 
equal temperature, it is colder and heavier than the air around it. 
When the outside force which caused it to rise is no longer effective, 
it sinks back to the surface. The air tends to return to its original 
position under such conditions, and is therefore said to be stable or 
in stable equilibrium. A case of inversion, in which warm air over- 
lies a surface layer of cold air, is evidently an example of marked 

Instead of assuming that the temperature is the same throughout 
or increases upward, let us assume that it falls with elevation, but 
that the rate of fall is less than the adiabatic rate, as shown in Fig. 
57. Any portion of this air having a vertical movement will change 
its temperature at the adiabatic rate. If started upward, it will be- 
come colder and heavier than the surrounding air and will settle 
back to its original position. If pushed downward, it will warm adia- 



batically and become lighter than the surrounding air. This action 
will, likewise, cause the displaced air to return to the initial posi- 
tion. Hence this air is stable, and any air is stable in which the lapse- 
rate curve slopes to the right of the adiabatic line that it would fol- 
low should it be displaced. The general rule may be stated thus: 
Unsaturated air is stable when its lapse rate is less than the adiabatic 
rate, and saturated air is stable when its lapse rate is less than the 
saturation adiabatic rate. Such air stays in position or returns to its 
position if forced out of it. Thermal convection is not possible so long 
as the air remains in this condition. 




10 o 10 



Fig. 57. Lapse Rate Showing a Stable Atmospheric Condition. 

Instability. If the lapse rate, as indicated by the heavy line, Fig. 
58, is greater than the adiabatic rate, air starting at any point and 
moving upward becomes progressively wanner than its surround- 
ings and therefore continues to rise indefinitely as long as the given 



lapse rate of the surrounding air continues. Likewise, air starting 
downward becomes progressively denser than the surrounding air 
and continues downward to the earth's surface. This is a case of in- 
stability; a little push in either direction sets the air moving in that 
direction. If air is heated near the surface and starts to rise, it will 
continue to rise as long as it is surrounded by air that is colder than 
itself. Air is unstable when its lapse rate is greater than the dry adia- 
batic rate. This condition is favorable to convection. If the lapse rate 
just equals the dry adiabatic rate, the air is in neutral equilibrium. 
For any lapse rate greater than the dry adiabatic rate, the air is un- 
stable but generally requires an impetus to start vertical movement. 
It does not start moving up or down automatically. 



- 800 





^ \ v 

x ' 


\ ^ v 


\ X 

\ X 


^ v x 




v * X 

x ^^^ 


\ * 

^ \ 

^ ^' 




\ * 

S N 

\ s 

X x \ 

X N 

+ ^ v x 




v x x 


\ X x 

\ N x 
\ X N 


k > ^- 

N \ 
^N^ \ 

N \ x \ 

S 'N 





^. . \ 


k > 

\ X 

\ X 

^ v v 

X \ 


\ ^ V 



\ ^ 

\ S N 


\ ^N 

\ * ^ v 

\ N v 

\ N \ 

\ \ 

x \ ^^ 

X >y 

s _ \ N 

x X. 

X ^. 

^^^ST~ *" 

x ^ x 

V N \ 


N x \ 


N \ 


*s^ N:S 

V v 

\ N 


X v^ 


20 .10 10 


Fig. 58. Lapse Rate of an Unstable Atmospheric Condition. 

The maximum possible lapse rate, except momentarily, is one in 
which the air gets cold so rapidly with height as to offset the tend- 


ency of expansion with decreasing pressure, and thus to give the 
air a constant density in the vertical. This effect requires a fall of 
temperature of 19F. per 1,000 feet (3.5C. per 100 m) of elevation, 
and is an extremely unstable condition. Any temporary increase of 
the lapse rate above this value would create a condition of instabil- 
ity that would result in automatic and almost explosive overturning 
of the air. It would require no impetus to start the movement; it 
would be auto-convective. Such a condition may occur momentarily 
in tornadoes and in a thin layer of air in contact with an intensely 
heated ground surface, when the air is unusually quiet. 

If the lapse rate changes in Fig. 58, as indicated by the heavy 
dashed line, the layer of air above the point of change is stable. Air 
from the unstable layer will descend until it meets the surface or 
rise until it reaches a point in the stable layer having a temperature 
identical with its own, and it will go no farther. Such a change in 
lapse rate may be occasioned by a warm current in the upper air, 
and such a current limits convection. Convection will stop at, or 
somewhat above, the bottom of the warm current, and if the air has 
not already been cooled to its dew point, there will be no conden- 
sation, no clouds, and no rain. Conversely, if there is a cold current 
in the upper air, convection will continue through it, and the rising 
air will probably cool below its dew point, resulting in cloudiness 
and rain. A stable condition of the atmosphere, therefore, favors fair 
weather; an unstable condition is conducive to cloudiness and rain. 

Conditional instability. The illustrations of stability and insta- 
bility assumed that the rising air did not become saturated. After 
condensation begins, rising air cools at the saturation adiabatic rate, 
and it is that rate rather than the dry adiabatic rate which then 
determines stability or instability. The condition of the atmosphere 
is at times such that the lapse rate may be represented by a line 
lying between the dry and wet adiabatic curves (Fig. 59). The lapse 
rate is less than the dry and greater than the wet adiabatic. Such 
air is therefore stable when unsaturated, and unstable when con- 
densation is occurring in it, and it is said to be in a state of condi- 
tional instability. Whether it is stable or unstable is conditioned by 
whether or not it has been cooled to saturation. 

In many cases, the cooling is adiabatic cooling due to lifting. As- 
sume a body of air having a considerable moisture content and a 
lapse rate intermediate between dry and wet adiabatic. If a portion 





Z 800 




10 10 



Fig. 59. Conditional Instability. 

of this conditionally unstable air is forced upward through the mass, 
it becomes definitely unstable, as illustrated in Fig. 59. As it is lifted 
and is cooled adiabatically, condensation begins at a height depend- 
ent on the moisture content and called the lifting condensation 
level If the air is forced farther upward, it cools at the saturation 
adiabatic rate, and becomes warmer than the air around it, and 
therefore unstable. The height at which instability begins .is the level 
of free convection. 

The instability of the rising air was latent in this case and was re- 
alized only when an outside force caused a lifting of the air to the 
level of free convection. Condensation in air forced to rise, there- 
fore, often makes thermal convection possible when it would not 
otherwise be so, and permits it to extend to greater heights than it 
otherwise would. When the air is definitely unstable, that is, when 
the lapse rate is greater than the dry adiabatic rate, condensation 



makes convection more active by increasing the difference between 
the temperature of the rising air and that of the surrounding air. 
Conditional instability is of frequent occurrence in the atmosphere, 
often in connection with widespread rain. 

Convective instability. Similarly, an entire layer of stable air may 
become unstable by lifting. The distribution of temperature and 
moisture must be such that the layer becomes saturated as it rises 
and that the bottom portion becomes saturated before the upper 
portion does (Fig. 60). The latter condition is frequently met be- 
cause the relative humidity often decreases rather rapidly from the 







.10 10 



Fig. 60. Instability of Lifting. The lifting condensation level is at A t and the level of 
free convection is at A 2 for air at the surface. 

surface upward. As such a layer rises, it expands under decreasing 
pressure, causing the upper portion to rise more than the lower. The 
upper portion therefore cools more than the lower, both because it 


cools longer at the dry adiabatic rate and because it cools through a 
greater distance. The lapse rate within the layer is thereby increased 
and becomes greater than the saturation adiabatic. Layers of this 
character, which, though originally stable, become unstable on ris- 
ing, are convectively unstable or in a condition of connective insta- 
bility (also called potential instability). A layer of air will be con- 
vectively unstable if the equivalent potential temperature decreases 
with height through the layer. 

Subsidence. In contrast, a subsiding layer of stable air has a de- 
creasing lapse rate and an increasing stability. As the base of the 
layer approaches the earth, the layer is compressed by the increas- 
ing pressure upon it. Hence, the top descends farther than the bot- 
tom and is warmed, adiabatically, more than the bottom. This re- 
sults in a smaller difference in temperature between top and bot- 
tom, and in a decreased lapse rate. Subsidence of a layer of stable 
air therefore increases its stability and may even cause a tempera- 
ture inversion. 

Turbulence in relation to lapse rates. The atmosphere, especially 
within a few hundred feet of the earth's surface, but to some extent 
at all elevations, is turbulent, and the many small irregular move- 
ments and eddies have vertical as well as horizontal components of 
motion. This is mechanical turbulence, produced by the action of 
the winds over the earth's surface. If the air is unstable, such turbu- 
lence starts convective currents which will then continue upward. 
The pilot recognizes these rising currents as "bumps." If the lapse 
rate is less than the adiabatic, the vertical movements started by 
turbulence are checked and damped. Hence turbulence is more pro- 
nounced in unstable air than in stable air. This is illustrated by the 
usual increase in the gustiness of wind by day over that at night. 
By day there is instability and convection, favoring gustiness and 
increased vertical movement; by night, the cooling of the surface 
layers favors stability and decreased moyement. 

The daytime turbulence slows down the heating process by bring- 
ing down cooler air. Whatever night turbulence there may be, owing 
to winds of more than 5 or 6 miles per hour, retards the night cool- 
ing by bringing down warmer air, and thereby tends to prevent 
frosts and to prevent or dissolve ground fogs. The so-called ther- 
mal?, often used by glider pilots to soar are rising currents of 


unstable air. Varying stages of instability are also responsible for 
much cloudiness and rainfall. For these reasons, the degree of sta- 
bility of the air is of practical significance. 

Atmospheric Layers 

Until recently it was supposed that the air above the first few 
miles grew thinner and colder by continuous gradation until it gradu- 
ally merged into outer space. No change in characteristics, except 
for the proportions of its constituent gases, was suspected. It was 
assumed that this high air had little influence on terrestrial affairs 
and presented few problems of scientific interest or practical concern. 
Among the most important advances in meteorology since the be- 
ginning of the twentieth century has been the discovery that the 
upper air has a complicated physical structure, with many theoreti- 
cal and practical bearings. In particular, upper-air exploration has 
shown that the atmosphere has some structural resemblance to a 
house of several stories; it is divided into layers, or strata, and each 
layer has its own peculiar features and behavior (Fig. 61), 

Stratosphere and troposphere. The first evidence of stratification 
in the upper air came with the discovery of what is now known as 
the stratosphere, which Sir Napier Shaw called "the most surprising 
discovery in the whole history of meteorology." It has been noted 
that the normal lapse rate is about the same in all parts of the world 
and beyond an elevation of about two miles becomes quite regular. 
It was natural to assume that this condition continued upward in- 
definitely, but the accumulation of data from sounding balloons en- 
abled Teisserenc do Bort and Assmann to demonstrate, between 
1899 and 1902, that the air ceases to become colder with elevation 
at a certain fairly sharp limit in the upper air, at an average eleva- 
tion of about 7 miles (11 km). From this surface upward for a dis- 
tance of a few miles, as has been confirmed by many subsequent 
observations, the temperature remains practically the same, or in- 
creases slightly; the lapse rate of the air at these elevations is zero 
or negative. The region is therefore nearly isothermal in a vertical 
plane, and was first known as the isothermal region. It is now called 
the stratosphere. It is evident that the air in this layer of the atmos- 
phere is in stable equilibrium; there can be no convection through it. 

The region between the earth and the stratosphere, where there 






















































Fig. 61. Layers of the Atmosphere. 

are frequent instability and convection currents, is known as the 
troposphere. The prefix tropo carries the meaning of a turning or 
overturning of the air, such as occurs in convectional movements, 
The boundary surface between the two regions, the level at which 
the troposphere ceases and the stratosphere begins, is the tropo- 
pause. Too little is known about the absorption and radiation of 
energy in the thin air at such levels to permit a complete physical 
examination of these atmospheric layers. 


Height of tropopause, and temperatures in the lower stratosphere. 

In the temperate latitudes of Europe, where records were first ob- 
tained and studied, the stratosphere was found to begin at a height 
of about 7 miles (11 km). With the accumulation of records from 
other parts of the world, it is now known that the height of the 
tropopause varies with latitude. The height is about 10.6 miles ( 17 
km) in equatorial regions, from which it gradually decreases toward 
the poles, both north and south, descending in polar regions to an 
elevation of only 4 or 5 miles (6-8 km), and possibly less. In addi- 
tion to this marked change in height with latitude, there are smaller 
changes related to the seasons and to barometric pressure at the 
surface. The tropopause is higher in summer than in winter and 
higher when the surface pressure is high than when it is low. Fig. 
56, sounding 1, which was obtained at about latitude 41 north, 
shows the beginning of the stratosphere at 34,000 feet, at a tempera- 
ture of 53C., and almost isothermal conditions above that level. 
In sounding 5, obtained at latitude 25 north, the stratosphere be- 
gins at 46,000 feet and at a temperature of 58 C.; above that height 
the temperature rises noticeably. 

Although vertical surfaces in the lower portion of the stratosphere 
are nearly isothermal, it is by no means true that the stratosphere is 
everywhere of the same temperature. The temperatures at the same 
elevation in different parts of the world vary widely. In equatorial 
regions, the normal lapse rate continues to a height of about 10 miles 
(16 km), until the temperature has fallen to -80 or -100F. (-62 
or 73C.). A temperature of 134F. was once registered at a 
height of 10 miles, above Batavia, Java. In polar regions, the tem- 
perature decreases to a height of only 4 or 5 miles ( 6-8 km ) above 
the earth and falls to -40F. or -50F (-40 or -45C.). In middle 
latitudes, the temperature at the tropopause, about 7 miles ( 11 km) 
above the surface, is about 60F. The higher the tropopause, the 
longer the lapse of temperature continues, and the lower is the tem- 
perature of the stratosphere. Hence, at heights of 5 miles ( 8 km ) or 
more, it is colder over the equator than over the poles. This is true 
in all seasons. There are movements of air in the stratosphere, per- 
haps the result of the temperature differences just mentioned; but in 
passing from the troposphere to the stratosphere, it has usually been 
found that the winds decrease in velocity fairly rapidly, without 
changing their direction. 


Ozone layer. Spectroscopic observations have shown that there 
exists in the atmosphere a total quantity of ozone which, if concen- 
trated at the surface of the earth under normal atmospheric pres- 
sure, would form a layer only one-eighth of an inch (3 mm) thick. 
The amount increases from equator to poles. It is greatest in spring 
and least in autumn. It occurs in greatest concentration in the layer 
between 20 and 37 miles ( 30 and 60 km ) above the earth, where it 
forms what is sometimes called the ozone layer. Some ozone occurs 
in the lower atmosphere, but its amount there is extremely small It 
is well known that ozone absorbs much more radiation than do the 
other permanent gases of the air, especially in the ultraviolet por- 
tion of the spectrum. Because of this absorption, temperatures to- 
ward the top of the ozone layer are much higher than in the strato- 
sphere below, and even higher than at the surface of the earth. (See 
Fig. 1.) 

The ozone layer acts as a filter, absorbing ultraviolet radiation. If 
it were not there, the full complement of ultraviolet reaching us 
from the sun would burn our skins, blind our eyes, and result in our 
destruction. But if the layer were thicker and absorbed all of the 
ultraviolet, we should also suffer, for some of this short-wave radia- 
tion is necessary to health and even to life. This slight and rarefied 
layer of ozone furnishes an excellent example of a nice adjustment 
of nature, an adjustment necessary to our life but entirely unsus- 
pected until recently. 

Ionized layers. At still greater heights than that of the ozone 
layer there are other interesting and significant strata in the atmos- 
phere. Information about these layers was first obtained through the 
development of long-distance radio communication, and exploration 
of the properties of the upper air has been continued by soundings 
made by instrument-carrying rockets. The layers are highly conduc- 
tive electrically and serve to turn certain radio waves back to the 
earth by refraction. The high electrical conductivity is due to the 
presence of ions, which are electrified, gaseous atoms, produced in 
the gases of the rarefied air by solar and cosmic radiation. 

By the accurate timing of radio waves of different lengths, and 
later by the use of rocket soundings, it has been shown that there 
are three separate ionized layers or regions in the upper atmosphere. 
These are referred to as the D, E, and F layers (Fig. 61). The low- 
est is the D region; it occurs at heights ranging from 37 to 60 miles 


(60 to 100 km) above the earth's surface, and turns back only the 
longest radio waves. It is not always clearly defined during the day 
and disappears completely after the sun sets. The next is the E, or 
Kennelly-Heaviside, region with an elevation ranging from 60 to 90 
miles (95 to 150 km) above the earth. This layer returns radio waves 
from 300 to 400 meters in length. The F region, also called the Ap- 
pleton layer, is divided into two parts, Fi and p2. The Fi layer in- 
cludes the region from about 90 to 150 miles ( 140 to 240 km) above 
the earth, and the 2 layer extends from about 150 to 220 miles (240 
to 350 km). These F layers return the short waves used in radio 
broadcasting, but some of the shortest, such as television waves, es- 
cape into outer space. Without these reflecting regions, long-dis- 
tance radio communication would be impossible. The effective 
heights of all the layers have daily, annual, and irregular variations. 

The name stratosphere has generally been used loosely to mean 
the entire extent of the atmosphere above the tropopause but is now 
more often confined to the layer between the troposphere and the 
ionized region of the Kennelly-Heaviside layer. The region above 
this boundary has been given the name of ionosphere. 

Temperatures in the upper atmosphere. The existence of warm 
layers in the upper air was inferred from physical studies of heat 
absorption and of the behavior of meteors and sound waves. The 
presence of two hot layers and one cold layer has been confirmed by 
records obtained by the use of rocket-borne thermometers. Tempera- 
tures begin to increase in the ozone region at a height of about 20 
miles (32 km) and continue to increase with height. This situation 
results in the formation of a hot layer between the 30- and 40-mile 
levels with a temperature of about 150F. Between the levels of 40 
and 50 miles, temperatures fall rapidly to form a cold layer of about 
30F. Above 50 miles (80 km), temperatures again increase to 
almost 2,000 F. at a height of 250 miles (400 km). This is in the 
ionized layer, and the high temperatures are due, in part at least, 
to the heat generated by ionization and to the absorption of solar 
radiation by cosmic dust. 


1. Assuming the air to have the average lapse rate and a surface tem- 
perature of 60F., if a certain mass of dry air at the surface is heated to 
72F., how high will it rise and what will be its temperature at that 


2. Assume that the maximum temperature on a quiet summer after- 
noon will be the temperature at which the air becomes unstable, as in- 
dicated by the temperatures at the surface and at 3,600 feet. 

(a) What will be the maximum temperature if the air at 3,600 feet 
has a temperature of 70F.? 

(b) If the temperature at 3,600 feet is 45F.? 

3. Let air having a temperature of 15C. at the surface of the earth 
rise 3 miles, with condensation occurring during the last mile of the rise. 
What is its potential temperature at the surface and after it has risen to 
3 miles, under the following conditions: 

(a) When the pressure is 1,000 millibars at the surface? 

(b) When the pressure is 1,000 millibars at 528 feet elevation? 

4. On a certain day, air has the following temperatures at the eleva- 
tions given: 50F. at the surface; 42F. at 1,000 feet; 45 at 2,000 feet; 
41 at 3,000 feet; 38 at 4,000 feet. 

(a) Make a chart of height against temperature and plot the lapse 

(b) What part of the air is stable? 

(c) What part is unstable? 

(d) When the surface air is heated it) , how high will it rise if no 
condensation occurs? 

5. On a September day at Drexel, Nebraska, the following upper-air 
data were obtained. ( Altitudes are expressed in meters and temperatures 
in degrees centigrade. The ground has an elevation of 396 meters. ) 

Altitude Temperature 

396 ... 8,7 

627 ... 13.9 

1,187 . . . . . 11.1 

2,443 -0.3 

3,094 . . -5.0 

3,292 . . . -7.0 

( a ) Plot the lapse rate. 

( b) What part of the air is stable? 

( c ) What part is conditionally unstable? 

6. Why is it more difficult for the surface temperature to fall below the 
freezing level after it has been raining than when it is dry? 

7. Explain the formation of an icicle. 



Because water is a large and essential constituent of living organ- 
isms, the earth is habitable only because of the large amount of 
moisture at its surface and in its atmosphere. The evaporation of 
water, forming a gas which mixes with the other gases of the air, 
and its condensation again at or above the earth's surface, are pro- 
cesses of the greatest practical as well as theoretical importance in 
the study of the weather. It has already been shown that the cooling 
of water vapor causes part of it to condense. In the atmosphere, cool- 
ing is the only cause of any significant amount of condensation. It 
is important to remember this fact in considering the causes of cloudi- 
ness and precipitation. 

In the long run, since the amount of moisture in the air is doubt- 
less becoming neither greater nor smaller, evaporation into the air 
must be balanced by condensation from the air. Over the earth as 
a whole, rainfall, plus dew, frost, and fog deposits, are equal to 
evaporation. This, of course, is not true, except by accident, for any 
one place on the earth or for any given period of time. The moisture 
evaporated is often carried great distances and held for long periods 
before being precipitated. 

Condensation on Solid Surfaces 

Condensation begins first on solid surfaces because these get 
colder than the general mass of air. The earth and all solid objects 
are better radiators of heat than is the air; at night they cool more 
rapidly than the air, this being especially true when the sky is clear 
and affords but little radiation itself. The air then loses some of its 
heat by radiation and conduction to the cold surfaces. 



Dew. Air that comes in contact with cold surfaces may thus be 
cooled below its dew point, in which case some of its moisture is 
condensed and deposited as dew on the cold objects. If the air is 
quite calm, the lower three or four feet may be appreciably cooled 
by conduction during a single night. Usually, only the air that comes 
in direct contact with the cold surfaces is cooled to its dew point. 
It is hardly correct, then, to say that dew falls; rather, it condenses 
where it is deposited. If the air is quiet, the cooling of the lower 
air produces an inversion of temperature, which decreases turbu- 
lence and so contributes to further stability and calmness. The air 
at the ground is thus left long in contact with the cold surfaces and 
is given a good opportunity to reach its dew point. By this process 
the air within a few inches of the ground may become considerably 
colder than that immeditaely above it. On the other hand, move- 
ment and turbulence in the lower air cause a mixing to a height 
of several feet. The cooling extends to a greater elevation than in 
quiet air, but, since more air is affected by the cooling process, it 
may be that none of it reaches its dew point. Hence, wind tends to 
prevent the formation of dew. 

Frost. When the dew point of the air is below 32F., moisture 
passes directly from the gaseous to the solid state. This process is 
called sublimation, and results in the formation of ice crystals, called 
frost or hoarfrost. Note that frost is not frozen dew. Frosts are classi- 
fied as light, heavy, or killing. A killing frost is defined as a frost 
that is destructive of the staple crops of the locality. Only the last 
killing frost in the spring and the first one in the fall are of special 
significance. The words frost, black frost, and dry freeze are some- 
times used in a broader sense to denote freezing weather unaccom- 
panied by hoarfrost. Light and heavy frosts are distinguished largely 
by the amount of the deposit and are without exact definition. 

Frosts occur most readily in low places, especially if there is no 
outlet. The cold, heavy air drains along the sloping surfaces into 
such low places and accumulates there, becoming still and stable and 
considerably colder than the general mass of air, thus creating a tem- 
perature inversion. In many parts of the world, fruit is grown suc- 
cessfully on slopes and in foothill regions, but not on adjacent val- 
ley floors, for this reason. Even on level ground, frosts may form 
when the general air temperature is well above freezing, especially 
if there is not sufficient wind to move and mix the air. On a cold win- 


ter day, frost often occurs on the inside of a window in a general 
air temperature of 70F. within the room, for the same reason that 
frost occurs outside, that is, by the loss of heat to a cold surface. An 
electric fan directed toward the window will clear it of frost by re- 
placing the cold air. 

The conditions necessary for the formation of dew or frost in na- 
ture are: ( 1 ) clear sky ( except that in cloudy winter weather a damp 
wind moving over cold ground may produce frost), (2) still, cool 
air in stable equilibrium, and (3) sufficient moisture to reach the 
dew point with a moderate amount of cooling. The prediction of 
frost takes account of these factors and of one further consideration 
in respect to the dew point. If the dew point is above 32F., con- 
densation will begin as dew, and the latent heat thus set free will 
retard the further cooling. Freezing temperatures are thus less likely 
to occur when the dew point is above freezing than when it is be- 
low 32F. The same process that causes dew and frost causes the 
sweating of cool objects on a hot summer day and the condensation 
of ice on cold pavements in winter. If the temperature falls below 
freezing but does not fall to the dew point, there will be a freeze but 
no deposit of frost. 

Protection against frost. In western and southwestern fruit- 
growing regions, most injurious spring frosts or freezes occur on 
"radiation nights" that is, under clear and quiet conditions, with 
cold air at the surface and an inversion layer not over 30 or 40 feet 
above the surface. Many citrus orchards in those regions, particu- 
larly in California, are protected from injury under such conditions 
by the use of small diesel-oil-burning heaters, giving much heat and 
little smoke. These are placed among the trees, sometimes as many 
as 60 to the acre (Fig. 62). By this means it is possible to raise the 
temperature of the greater part of the grove by as much as 12F. 

In the use of these heaters, three factors are effective in prevent- 
ing injury: (1) The lower air is warmed by the heat produced; (2) 
the fires create small convection currents which mix the air to about 
the heights of the tree tops; (3) such smoke as is formed acts as a 
blanket to retard cooling by radiation. But smoke is avoided as much 
as possible because it is a public nuisance and because, when it re- 
mains in the air by day, it retards surface warming. Larger fires 
would be less effective, by reason of carrying the heat above the 



trees, as well as being dangerous to the trees. It is evident that such 
protection is not feasible where freezing temperatures occur with 
cold winds and not as a result of radiation. 

Fig. 62. Oicliaul 1 1<. itcrs in a Yoimtf California (Jtius (iiovc. S 

Following World War II, the availability of war-surplus airplane 
engines and propellers led to the development of wind machines like 
the one in Fig. 63. Such machines do not heat the air, but they do 
stir the low inverted layer, forcing it to mix with warmer and lighter 
air from above. When properly operated, wind machines will usually 
prevent frost with no smoke and with less inconvenience than is as- 
sociated with orchard heaters. 

No considerable part of continental United States is entirely im- 
mune from frost, but in the southern half of Florida, certain limited 
areas in California and Arizona, and a small area in southern Texas, 
frosts are sufficiently rare to permit the growth of citrus fruits and 
winter vegetables, but not without occasional losses. The Hawaiian 
Islands are entirely free of frost at elevations below 2,500 feet, and 
Puerto Rico is also free of frosts. 



Fig, 63. Wind Machine for Fighting Freezes. This type of device is being exten- 
sively used in the California citrus groves. Most freezes are associated with sharp 
temperature inversions and may be averted by thoroughly stirring the lower air. 
Sunkist Photo. 

Condensation Above the Earth's Surface 

Any visible atmospheric phenomenon, except clouds, that depends 
on the presence of water in the air is called a hydromctcor. 1 Hydro- 

1 This description agrees with the United States Weather Bureau definition but does 
not agree, with regard to fog and haze, with the definition of the International Me- 
teorological Organisation. See Weather Bureau Publication No. 1445, Weather Glos- 
sary, Washington, D.C., Superintendent of Documents, 1946, p. 155; and Sverre 
Petterssen, Weather Analysis and Forecasting, New York, McGraw-Hill Book Com- 
pany, 1940, p. 37. 


meteors are mainly the result of the condensation or sublimation 
of the water vapor in the air. Condensation in the free air, as at the 
surface, is the result of cooling. Condensed moisture takes many 
forms, because of variations in the moisture content of the air, in its 
movements and turbulence, and especially in its temperature and 
its rate of cooling. These forms have been described in detail and 
given specific names and symbols by an International Meteorological 
Committee. The same symbols are in use on weather maps in all 
the principal countries of the world. The more important of the con- 
densation forms are discussed in this chapter. 

Nuclei of condensation. If air is perfectly free from dust, it may 
be cooled below its dew point without any condensation; the air is 
then supersaturated. Moreover, ordinary mineral dust, as from a 
land surface, may be added to such supersaturated air without start- 
ing condensation. But if smoke or salt spray from the ocean is added, 
rapid condensation occurs. In fact, with such substances in the air, 
moisture will begin to condense before" 100 per cent humidity is 
reached. Some of the ocean salts and some of the products of com- 
bustion have the quality of absorbing moisture from the air and for 
this reason are said to be hygroscopic. Apparently the presence of 
hygroscopic, or at least water-soluble, particles is essential to the 
condensation of moisture in the air in important amounts. Such 
particles are called nuclei of condensation. This term refers to par- 
ticles of microscopic size, not to the visible dust or smoke particles 
in the air. Fires, ocean spray, explosive volcanoes, and burning me- 
teors furnish large numbers of hygroscopic nuclei. Tests that have 
been made show that condensation nuclei are usually present in the 
air in adequate numbers. 

Fog, haze, and drizzle. Fog may be defined as almost microscopi- 
cally small drops of water condensed from and suspended in the air 
near the surface of the earth in sufficient number to reduce the hori- 
zontal visibility to 0.6 mile or less. Fog may also be defined briefly as 
stratus cloud near the earth's surface and enveloping the observer. 
Fog particles vary in diameter from about one-tenth to one-hun- 
dredth of a millimeter. Droplets of all these sizes often occur in the 
same fog. They frequently occur in a supercooled liquid state at 
temperatures much below freezing, even as low as 20F. Such 
supercooled fogs produce a rapid icing of aircraft moving through 


Accumulations of dust or smoke in the air are sometimes called 
dust fogs or smoke fogs, but they should be distinguished from true 
fogs. Smoke furnishes numerous hygroscopic nuclei and probably 
facilitates the formation of fog. Certainly smoke darkens fog and re- 
duces the visibility. For this reason, thick fogs are more frequent in 
smoky cities than in adjoining country districts, but smoke abate- 
ment, though very desirable in itself, would not put a stop to fogs. 
The blend of smoke and fog is called smog. 

Fogs are now classified in four densities in terms of their effect 
on visibility, as follows: 

Light fog visibility % mile or more. 
Moderate fog visibility between %6 and % mile. 
Thick fog visibility between % and % G mile. 
Dense fog visibility less than % mile. 

Fogs merge gradually into drizzle as the droplets become larger. 
Drizzle implies light rain, which is falling, or at least can be felt on 
the face. On the other hand, when the fog droplets become smaller 
and less numerous, fogs grade into moist haze. In haze there is no 
visible obscuration of near-by objects, within about half a mile, but 
distant objects become blurred and the sky has a gray appearance. 
Nearly the same effect may be produced by dry haze, resulting from 
dust or smoke or from optical irregularities of th air. Four import- 
tant processes by which the saturation necessary to produce fogs is 
obtained are discussed in the following sections. 

Radiation fogs. The loss of heat by radiation often results in the 
saturation of the lower air and the development of a fog. Such fogs 
are named radiation fogs. They are of two types: ground fogs and 
high-inversion fogs. Ground fogs are a result of the cooling of the 
earth's surface and the lower air at night, producing an inversion of 
temperature, which prevents convection and reduces turbulence. 
They occur principally in the early morning hours. Sometimes only 
dew or frost follows such cooling, but at other times the entire mass 
of air to a height of a few feet or a few hundred feet is cooled below 
its dew point, and then there is fog. 

A light wind of 4 or 5 miles per hour is sufficient to produce tur- 
bulence when moving over uneven ground or around trees and 
buildings, and such a wind is conducive to fog; but higher winds 


carry away the cooled air, destroy the inversion, and prevent fog, 
Fogs of this character do not extend to any considerable height, 
frequently not over 100 feet; hence, their name, ground fags. Be- 
cause clear weather permits rapid cooling, ground fogs are fair- 
weather fogs; that is, the air is bright and clear above and also at 
the surface when the sun breaks through the "vapors that did seem 
to strangle him/' But they are most likely to occur when a clear night 
follows a cloudy day, because then the surface and the lower air 
start the night cool. 

High-inversion fogs. During the winter season, it often happens 
in certain regions that cool, quiet, moist air overlies the earth and is 
itself overlain by warmer, drier air at elevations from 300 to 2,000 
feet. This occurrence prevents upward movement by convection or 
turbulence and facilities cooling by radiation. When this situation 
persists, the continual cooling of the lower air, night after night, re- 
sults in a sharp inversion at the boundary of the two layers and the 
formation at that boundary of a high-inversion fog, also called in- 
version fog. This is really a low stratus cloud. The further cooling 
of the already cool air by radiation at night often causes the con- 
densation to build downward from the cloud to the earth, causing 
a dense surface fog at night. When air from polar regions, moves 
over the North Atlantic and becomes stagnant over Europe, such 
fogs frequently form and persist day and night for several days or 
even weeks. In this country, in valleys near the seacoast, especially 
in southern and central California, there are frequent inversion fogs 
during the winter months. They are especially prevalent and dense 
in the San Joaquin Valley of California. 

Advection fogs. A second important process by which fogs are 
formed is the movement of warm, moist air over a cold surface. 
These are called advection fogs. The first essential in the formation 
of such a fog is the importation of warm, moist air. The second is 
the cooling of the air to saturation by its movement over the cold 
surface. Third, turbulent mixing extends this saturated layer to con- 
siderable heights. Such fogs may occur with moderately strong 
winds, and the higher the wind, the deeper the fog layer will be, 
if formed at all. They often occur, also, with cloudy weather and 
either by day or at night. They are often dense, reducing ceiling 
and visibility to zero, and they dissipate slowly. 

Over continental interiors advection fogs are more frequent in 


winter, when the ground is cold or snow-covered. On western sea- 
coasts in temperate latitudes, warm, moist, sea air, drifting inland 
over radiation-cooled land, is often the occasion for such fogs. At 
sea they occur where there are adjacent bodies of water of contrast- 
ing temperatures. The dense and persistent fogs in the vicinity of 
Newfoundland are of this character and result from the movement 
of air from the warm water of the Gulf Stream to the cold water of 
the Labrador Current. Fogs of this type are also frequent from 
Greenland eastward to Iceland and Spitzbergen, owing to the meet- 
ing of relatively warm and cold ocean waters in this region. 

Evaporation fogs. When cold air moves over warm water, it 
often happens that the moisture evaporated from the water and 
added to the cold air is sufficient to produce saturation and start 
condensation. This results in a fog, beginning at the water's surface 
and building upward. It has the appearance of steam rising from 
the water. Hence such fogs are sometimes called steam fogs. The 
same process is observed when a pan of warm water is placed in 
cold water. It will be noted that this process involves the advection 
of cold air, and for that reason these are often classified as advection 
fogs. They may become dense at the surface to a depth of 50 to 100 
feet, but they do not extend to any great height. They occur over 
rivers, lakes, and oceans, especially in Arctic regions where there is 
open water overlain by air many degrees below freezing. 

Another type of evaporation fog often occurs when rain falls from 
a warm layer of air through an underlying layer of cold air. Evapora- 
tion from the falling warm rain may saturate the cold air, if the 
temperature of the raindrops is higher than the dew point of the air. 
If this air is unstable, or if the wind is moderate or stronger, the 
moisture is carried upward to form a low stratus or stratocumulus 
cloud deck. If the air is stable and the wind light, the condensed 
moisture remains near the surface, forming a fog. Fogs of this type 
occur near the boundary between two masses of air of different tem- 
peratures. Such a boundary is called a front, and the fogs are known 
as frontal fogs. A distinction is sometimes made between prefrontal 
and postfrontal fogs. 

Upslope fogs. When air moves upslope against a mountain side, 
or even up a gradually sloping plain, the adiabatic cooling due to 
ascent may result in saturation and the development of an upslope 
fog. The air must have a rather high relative humidity to begin with 


and must be stable. If there is convective instability, clouds will 
form but no fog. Radiation cooling of the air at the surface is often 
a contributing factor in the development of these fogs. 

Fog cost and dispersal. Dense fogs are very expensive affairs. 
They are the cause of many accidents; they delay traffic by land 
and sea, and cause many shipwrecks; they increase cleaning bills and 
the use of gas and electricity, and are a special menace to aviation. 
London "pea soup" fogs are perhaps the densest and blackest in the 
world, and there, when the fog is dense, physicians cannot answer 
calls, mail is not collected, and the fire apparatus goes to a fire at a 
snail's pace. Methods developed for dissipating fog over airplane 
landing fields, or other small areas, are based on warming the air 
above its dew point. This is analogous to the use of orchard heaters 
to prevent freezing. It is easier under quiet atmospheric conditions, 
as in radiation fogs, and more difficult in connection with advection 
fogs. Even in a gentle breeze the cleared air is soon replaced by 
foggy air, and it becomes necessary to warm a considerable volume 
of air quickly, if the field is to be kept cleared. 

Clouds and Precipitation 

In contrast to fogs, which result from cooling by conduction or 
radiation, clouds are chiefly the result of the dynamic cooling pro- 
duced by expansion under reduced pressure. By far the most impor- 
tant cause of clouds is the adiabatic cooling resulting from upward 
movement of the air. Some clouds are formed by the mixing of 
warmer and cooler air. 

The exact process or processes by which the minute cloud drop- 
lets grow to sufficient size to fall to the earth as precipitation are 
still not fully known. One theory that has been widely publicized is 
the Bergeron ice-crystal theory. This theory assumes that at least the 
tops of all clouds from which appreciable rain falls are at tempera- 
tures below freezing and consist of both ice crystals and supercooled 
water droplets. The latter are drops of water that were condensed 
to the liquid form at temperatures above freezing and have remained 
liquid after being cooled to temperatures below freezing, sometimes 
much below. There is no complete physical explanation of this phe- 
nomenon, but it is of frequent occurrence in the atmosphere, as 
aviation experience has proved. The fact on which the Bergeron 


theory is based is that the saturation vapor pressure over super- 
cooled water drops is greater than that over ice. When the two exist 
together in a cloud, there will be evaporation from the droplet and 
condensation on the ice crystal. It is thus that the crystals grow to 
sufficient size to fall. As they fall through the cloud, they may grow 
in size by further condensation and by coalescence with other drops 
after they melt. 

Observations have tended to support this theory in the main, but 
rain does sometimes fall from clouds in which the temperatures are 
all above freezing, especially in the tropics. To explain these cases, 
other processes are suggested as probably effective in the growth of 
small raindrops. Some of these are: (1) the presence of unusually 
hygroscopic nuclei; (2) a vapor pressure gradient due to the pres- 
ence of drops of differing temperature; and (3) the coalescence of 
drops of differing size as they collide in the turbulent air. 

Rain falls beneath the air in which it is formed, or it is carried 
short distances by the wind. Even if the space above us to the top 
of the air were saturated, it would not as a rule contain enough 
water to make more than an inch of rain. Such a condition of com- 
plete saturation never occurs, and moreover, only a small part of the 
moisture in the air is ever removed by natural processes of conden- 

It is evident, therefore, that a large amount of water cannot fall 
from a given mass of air, but can come only from a continual re- 
newal of the moisture supply. Hence, one necessary condition for a 
heavy rain is a continuous supply of moist, rising, inflowing air. Rain 
may be held aloft by rapidly rising air for a time, and then suddenly 
be released when the updraft ceases; this results in an extremely heavy 
rain of short duration and over a small area. Such a shower may be 
called a cloudburst. The word should be confined to rain of this 
character but is sometimes erroneously applied to any heavy rain 
in mountain regions, where the run-off from a large area is collected 
into narrow valleys, giving the appearance of rain heavier than has 
actually fallen. 

The magnitude of the operations involved in the production of 
rain is seldom appreciated. One inch of rain weighs 113 tons per 
acre or 72,300 tons per square mile. A general rain of one inch over 
the state of North Dakota means the precipitation of 5 billion tons 
of water. All this water has first been lifted high into the air. The 



tremendous energy of the natural forces involved and the futility 
of trying to control them or to produce important amounts of rain- 
fall artificially are evident. Nature produces the necessary sustained 
upward movement in one of the ways mentioned in the next three 

Penetrative convection. The uplift may be by means of local con- 
vection currents, when certain portions of the lower air become so 
much heated that they are able to penetrate the overlying air. In 
such cases, descending columns of air are to be expected between 
the rising columns, as indicated in Fig. 64. This penetrative convec- 
tion is to be distinguished from the general expansion and upward 
movement of an entire, extensive layer of air. In cases of penetra- 
tive convection, clouds are likely to occur in relatively small masses, 
such as detached, flat-base cumuli. After condensation begins, the 
retarded cooling favors rising of the air and increased thickness of 
the cloud layer, and thus cumulonimbus clouds and thimdershowers 
frequently occur. Convective rainfall is therefore usually in the form 
of heavy showers of short durationinstability showers. 

On quiet summer afternoons when such cumuli are forming by 
thermal convection, we may reasonably assume that the air is of the 
same character from the ground level to the cloud bases. With that 
assumption, the condensation level, that is, the height of the clouds, 
can be calculated roughly, given the temperature and dew point 
at the surface. It is to be noted not only that the temperature of 
the rising air falls at the adiabatic rate (5.5F. per 1,000 ft.), but 
also that its dew point falls. The rate of fall of the dew point varies 

Fig. 64. Penetrative Convection Producing Summer Cumuli. 


with different humidities of the air. An approximation often used 
is 1.1F. per 1,000 feet. The rising vapor expands because of re- 
duced pressure, and this increase in volume decreases the concen- 
tration of the vapor and, hence, lowers the dew point. Calling the 
height of the cloud bases in feet, H; the temperature at the surface, 
Toy the dew point at the surface, D ; the temperature at the bases of 
the clouds, Tu; and the dew point there, Dh, we have 

T h = To -( 5.5/1,000 )H, 
and we also have 

D h ^ Do -( 1.1/1,000) H. 

Because H is the height at which condensation is beginning, Th 
and Dh represent the same temperature. Hence, 

To- (5.5/1,000) // == D- (1.1/1,000) H, 
from which 

4.4 H = 1,000 (To -Do), or H= 227 (To- D ). 

The values of T and D may be obtained by observation, and then 
the height of the cloud bases is easily calculated from this equation. 
This height is the convective condensation level. 

Orographic uplift Air may be forced upward by the movement 
of winds over rising ground. When winds move across a mountain 
range, large masses of air are made to rise. Continuous sheets of 
cloud, of the stratus type with flat bases, often result, and continu- 
ous rain also. As the air moves downward on the other side of the 
mountains, it is dynamically warmed and becomes dry and clear. 
Where winds are prevailingly from one direction across a mountain 
system, the windward side is wet and the lee side dry. The Sierra 
Nevada and the Rocky Mountains are wet on their western slopes 
and dry on their eastern. The Hawaiian Islands, with an elevated 
central backbone, have a very wet side facing the persistent north- 
east trade winds, and an opposite very dry side. The uplift due di- 
rectly to the slope of the terrain may not be sufficient of itself to 


cause rain, but if the air is conditionally or convectively unstable, 
even a moderate upslope movement may be sufficient to start con- 
vection and result in heavy precipitation. This probably accounts 
for much of the heavy rainfall in some mountainous regions, 

Convergence and eddy motion. When winds from different di- 
rections converge toward a center, as is the case in some of the 
storms to be studied later, some of the air is forced up, often result- 
ing in clouds and precipitation. Also, when currents of air of differ- 
ing temperatures meet at an angle, the heavier air will remain in 
the lower position, and the lighter air will be forced to rise. In both 
these cases the air is said to converge, and such convergence is the 
chief cause of cloudiness and precipitation, outside the tropics and 
mountain regions. In stable air, convergence is often attended by 
continuous cloud sheets and steady, prolonged rain, because the 
warm air moves slowly up an inclined plane rather than vertically 
upward, as in convective currents. In other circumstances, conver- 
gence furnishes the impetus to the ascent; of convectively unstable 
air, as in the case of orographic uplift, and causes cumulonimbus 
clouds and showers. 

The upper layer of the atmosphere may have a turbulent wave 
motion because of a difference in density of air masses moving from 
different sources. The air at the tops of these waves may be cooled 
below its dew point, while the lower portions of the waves remain 
unsaturated. Under these conditions, clouds form in long lines or 
rows, such as are frequently seen in cirrocumulus, altocumulus, and 
stratocumulus types. Rain from this wave-like movement is not to 
be expected. 

Forms of Precipitation 

As the drops of water or particles of ice which clouds are com- 
posed increase in size, they begin to fall more rapidly and eventu- 
ally reach the ground as precipitation, unless held up by ascending 
air currents or evaporated on the way down. Precipitation takes var- 
ious forms, depending upon the temperature at which condensation 
takes place and the conditions encountered as the particles pass 
through the air. 

Rain. The words rain and rainfall are often used to include all 
forms of precipitation, but in this paragraph rain refers specifically 




Fig. 65. Snow Crystals. Microphotograplm by \V. A. Bentley, courtesy, V. S. Weather 


to moisture which falls to the earth in a liquid state. Raindrops vary 
in diameter from 0.004 inch, in mist, to 0.2 inch in thunderstorm 
rain. There is a natural limit to the size of raindrops. Large drops 
falling through quiet air break up into smaller ones when they at- 
tain a \elot it\ of 18 miles per hour. Conversely, no rain can fall 
through an ascending current of this velocity. 


Snow. Snow is formed by the crystallization (sublimation) of 
water vapor at temperatures below freezing. Snowflakes are crystals 
of many beautiful, lacy patterns (Fig. 65). The fundamental form 
is hexagonal, but this is subject to much intricate elaboration, ap- 
parently influenced by the temperature, and perhaps also by the 
rapidity of condensation. Large snowflakes are formed by the com- 
bination of many small crystals, usually at temperatures not much 
below freezing, never at very low temperatures. At very low tem- 
peratures there can be but little moisture in the air, and therefore, 
under such conditions, precipitation is likely to be light, but it is 
never "too cold to snow." Snow has been recorded in Alaska at a 
temperature of 52F. As previously noted, a snow cover, being a 
poor conductor, keeps the soil temperature higher than it would 
otherwise be under winter conditions, but it keeps the air tempera- 
ture lower, because it is not much warmed by sunshine and cools 
rapidly by night. A snow cover is of much agricultural value in re- 
gions where the winters are severe. It prevents the soil's freezing as 
deeply as it otherwise would, and thus protects the roots of plants. 
Snow that accumulates in mountain regions during the winter and 
gradually melts in spring and summer is of great economic value in 
affording water supplies and maintaining the flow of rivers. On the 
other hand, the removal of snow from streets, roads, and railroads 
involves a large annual expense, in regions where the snowfall is 

Hail. Hail consists of hard, rounded pellets of ice, or of ice and 
compact snow. When a hailstone is cut in half, it is seen to be com- 
posed of concentric layers of differing densities and opacities ( Fig. 
66). Hailstones as large as marbles are common, and sometimes 
stones of much greater size occur. At Potter, Nebraska, on July 6, 
1928, a few very large stones fell, one of which was 5 inches in diam- 
eter and weighed 1% pounds. 2 Large flattened disks of ice are 
sometimes found; these are composed of several stones, formed in- 
dependently, and frozen together while falling. The destructive ef- 
fects of heavy hail, especially in the beating down of growing crops 
and the breaking of glass, are great. The area of destruction in any 
one storm is usually small, although occasionally quite extensive. 
Hailstorms are frequent in the central valleys of the United States, 

2 T. A. Blair, "Hailstones of Great Size at Potter, Nebraska," Monthly Weather Re- 
view, Aug. 1928, Vol. 56, p. 313. 


and and In that 


of hail Bail falls in a 

and In is an 

of air. In this air, 

as bat the of 

are by the 

are the is 

66, Typical l f S. 

a of ice by 

a to 

of the a of 

a of the en- 

counter and be 

the ice In 

by the of the 

is the of the of 

is in 


It Is not to 

and are 


formed simply by falling through subfreezing layers of air and cap- 
turing the supercooled water drops with which they collide. Hail- 
stones sometimes acquire several alternate layers of clear and opaque 
ice and reach a large size before falling to the ground. The ulti- 
mate size of a hailstone appears to depend mainly upon the upward 
velocity of the air, the concentration of supercooled water in the air 
through which it moves, and the length of its path through such air. 

Snow grains, sleet, and glaze. Small grains of snowlike structure, 
forming opaque white pellets, are known as snow grains. Sometimes 
there is a fall of even smaller and flattened grains, consisting mostly 
of ice needles. This is called granular snow. At other times the grains 
are larger, rounded, more crisp, and rebound when striking hard 
ground, and are then called soft hail. Sleet, as the term is now offi- 
cially used in America, means precipitation in the form of small 
particles of clear ice which are originally formed as raindrops and 
are later frozen as they fall through a layer of cold air. In Great 
Britain, and sometimes popularly in this country, the word desig- 
nates a mixture of rain and snow or partly melted snow. 

Precipitation sometimes occurs in the form of rain composed of 
supercooled drops which freeze rapidly upon striking either horizon- 
tal or vertical surfaces. This results in the formation of a coating 
of ice on trees, wires, paving, and other objects. Such a deposit is 
called glaze. Its occurrence is often popularly called an ice storm. 
The damage to trees and wires, resulting from breakage by over- 
weighting, is often large, especially when the storm is followed by 
high winds (Fig. 67). Deposits more than 2 inches in diameter have 
often been observed on wires and twigs. The slippery condition pro- 
duced on paved walks and roads creates a serious hazard to pedes- 
trians and motorists. There will be rapid and heavy icing of aircraft 
in flight in such a storm. 

Artificial Rain Stimulation 

From the foregoing discussion, it is obvious that much remains to 
be learned about the physical processes involved in condensation 
and precipitation. Some very interesting experiments, however, con- 
ducted by Langmuir and Schaefer in 1946, have stimulated many 
scientists to renew the search for the secrets of the raindrop. 

Operating on the theory that clouds sometimes fail to release 



Fig. 67. Telephone Lines and Trees Broken by the Weight of Glaze Ice Near 
Mount Freedom, New Jersey, January, 1953. Damage to communication lines be- 
comes severe when the diameter of ice covering the \\ires exceeds % inch. The Bell 
Telephone System suffered damage's of $8,000,000 from a single ice storm in January, 
1951. Courtesy, Southwestern Bell Telephone Company. 

their moisture because of lack of adequate condensation nuclei ( see 
page 7), several commercial companies in the United States have 
seeded clouds over large areas with silver iodide crystals. These 
crystals are known to be good hygroscopic nuclei. They are seeded 
into the cloud from airplanes or by generators located on the ground 
(Fig. 68). The seeding operation cannot create a rain situation at 
will, but may be able to stimulate more rain from a favorable situ- 
ation than would fall normally. 

On the other hand, rain-stimulating contracts are usually made 
for areas that are experiencing abnormally dry weather. Climato- 
logically, these areas should receive more rain in the near future 
than they have received in the recent past. In other words, the law 
of averages seems to favor getting more rain regardless of whether 
or not the clouds are "stimulated." Whatever the merits of rain 
stimulation, it is almost certain to bring about a better meteorologi- 
cal understanding of the processes of condensation and precipita- 



Fig. 68. Cloud Seeding Generator Foundry coke is impregnated with a carefully 
controlled solution of silver iodide and placed in a ]mpp< r A sc rrw-type feed releases 
fuel into the furnace at lower right. Silver iodide i,s vapon/cd in the intense heat of 
the burning coke and is emitted into the air stream at the rate of about 1,000,000,000,- 
000,000 (10 15 ) crystals per minute, each crystal being the potential center of a rain- 
drop. (Note vents at top of generator.) Courtesy, Water Resources Development 
Corp., Denver, Colorado. 



1. On a calm, clear spring evening, the temperature begins to fall after 
5 P.M. at the rate of 2F. per hour until the dew point is reached, and 
thereafter at 1F. per hour until 5 A.M. Assuming these basic conditions, 
answer the following questions, using the 5 P.M. data listed below, (a) 
When will condensation begin? (b) Will it be dew or frost? (c) What 
will be the minimum temperature? 

(1) Temperature, 50F.; Relative Humidity, 61 per cent. 

(2) Temperature, 50F.; Relative Humidity, 38 per cent. 

(3) Temperature, 50F.; Relative Humidity, 80 per cent. 

(4) Temperature, 45F.; Relative Humidity, 57 per cent. 

2. At a noon observation on a quiet summer day, the temperature of 
the air is 90, the wet-bulb thermometer reads 67, and detached cumu- 
lus clouds are observed: 

(a) How high are the bases of the clouds? 

(b) What is the temperature at the bases? 

(c) If the clouds are 1,200 feet thick, what is the temperature at 
the top? 

3. If air on a plain 3,000 feet above sea level, having a temperature 
of 42 and a dew point of 36, is forced over a mountain at an elevation 
of 12,000 feet above sea level and then descends on the other side: 

(a) At what height will condensation begin? 

(b) What will be the temperature at the mountain top? 

(c) What will be the temperature when the air has descended on 
the other side to its original altitude of 3,000 feet? Note: At this 
elevation and temperature, consider the wet adiabatic rate to 
be 3F. per 1,000 feet. 




In the discussion of convection, we have noted some of the rela- 
tions between the pressure, temperature, and movement of the air, 
with particular attention to vertical movements. Additional rela- 
tions between these elements of the weather are now to be noted, 
especially with reference to horizontal, or approximately horizontal, 
movements of the air. 

Pressure Gradients 

Large numbers of pressure records have been accumulated dur- 
ing the past hundred years, from all parts of the world, and they 
show that the pressure of the air is variable in a number of different 
ways. First, there is continuous variability of pressure at the same 
place from hour to hour. Second, pressures differ in adjacent places 
at the same time. Third, average pressures in different parts of the 
world are not the same. And finally, average pressures at a given 
place change with the change of season; they are not the same in 
winter as in summer. Pressure differences result from vertical and 
horizontal movements of the air, brought about by differences in 
density, and these, in turn, are due chiefly to temperature differ- 

Isobars and pressure gradients. To represent the various pres- 
sures over an area, lines known as isobars are drawn on a map 
through points of equal pressure at a chosen level. They may repre- 
sent the distribution of pressure at a definite time, or the average 
distribution for a given period. On surface weather charts, isobars 



are usually drawn for each 3-mb variation in pressure (Fig. 69). 
Isobars may also be drawn on vertical cross sections through the at- 
mosphere showing the distribution of pressure with height (Fig. 
70). In the three-dimensional atmosphere, isobars become isobaric 
surfaces. If one could connect all the points over California having 
exactly 800-mb pressure, they would form a surface rather than a 
line; but where the 800-mb surface was penetrated by mountains, 
there would be an 800-mb isobar along the mountain slope. 

The curved lines in Fig. 69 represent isobars on a map, showing 
the pressure decreasing from 1020 mb at the left to 1011 mb at the 
right. Air pressure, as measured by the barometer and as represented 
by isobars, is a force proportional to the weight of the air above 
the point of measurement, Simultaneous differences of pressure over 
an area, therefore, cause movements of air tending to equalize the 
pressure. A force pushes the air from the region of higher baro- 
metric pressure toward the region of lower pressure. 

The difference between the pressure at the points a and b is the 
force that is pushing the air at a toward b. In every case, the mag- 
nitude of the force depends on the difference of pressure, that is, 
on the rate of change of pressure with distance. The rate of change 
of pressure per unit horizontal distance is called the pressure gradi- 
ent. It usually means the change in a direction perpendicular to the 
isobars, since that is the direction in which the change is most rapid; 

but note that there are components of 
this gradient in other directions, also, 
such as cd. The gradient is expressed in 
millibars per hundred miles, per hun- 
dred kilometers, or per degree of lati- 



tude. Since the force increases as the 

1020 10)7 io*i4 10*1 1 gradient gets larger, the rate of move- 
!? Q T K A D ment of the air also increases. Both the 

Fig. 69. Isobars and Pressure 

Gradient. direction and the speed of the wind are, 

therefore, the result of the pressure 

gradient, but the actual movement of the air is modified by the 
earth's rotation, by centrifugal force, and by friction. 

Isobaric surfaces. The vertical distribution of pressure in the air 
above a given area may be represented by lines drawn to indicate 
the heights at which pressures are equal. Such lines then represent 
isobaric surfaces in the atmosphere. Just as the ground-level pres- 



Fig. 70. Vertical Cross Section of Isobaric Surfaces and Resulting Air Movement. 

sures are not equal over the earth, the isobaric surfaces above the 
earth are not, in general, parallel with the ground but are warped 
in various ways. Although isobaric surfaces are not necessarily par- 
allel to each other, they can never intersect because one point can- 
not have two pressures at the same time. If the points A, B, C, and D 
in Fig. 70 have pressures ranging from 1,023 millibars at A to 1,014 
millibars at D, then the isobaric surfaces above ED may be as rep- 
resented in the figure. The intersections of these surfaces with the 
ground are isobars. Consider the horizontal line FG at some dis- 
tance above the earth, and note that the pressure is greater at F 
than at such points as M and N, although the latter are nearer the 
ground. There is, therefore, a pressure gradient outward and down- 
ward from F, and the air flows out from the region of higher pres- 
sure as indicated by the arrows. 

Gradient Winds and Surface Winds 

A horizontal pressure gradient, if it acted alone, would lead to 
the flow of air along the direction of the gradient or perpendicular 
to the isobars. Motion in the atmosphere under a pressure gradient, 
however, is profoundly modified by an effect due to the rotation of 
the earth, and the result is a flow perpendicular to the gradient in- 
stead of along the gradient, except for a greater or less deviation 
produced by friction in the lower levels. 

Effect of the earth's rotation. An object moving in any direction 
over the surface of the earth tends continually to turn toward the 
right in the Northern Hemisphere and toward the left in the South- 
ern Hemisphere. This deflection is the effect of the rotational mo- 
tion of the earth and the movement of the body relative to the sur- 
face of the earth. The effect is the same as if the earth were at rest 
and a force were acting on the moving body. This influence is known 
as the deflecting force of the earth's rotation, or the Coriolis force. 



Assume that a long-range cannon is located just north of the equa- 
tor and is aimed due north at a target located on the 60th parallel. 
When the cannon is fired, the projectile will not travel due north, 
but will follow a trajectory to the right of north as indicated in Fig. 
71. This occurrence is due to the relative difference in eastward 
velocity of points on the respective parallels. In like manner it may 
be shown that, regardless of direction of movement, the deflection 
is always to the right in the Northern Hemisphere and to the left in 
the Southern Hemisphere. 


Fig, 71. Effect of the Coriolis Force. 

The Coriolis force is not a real force, but only an apparent one, 
because we determine direction of motion relative to the earth's 
surface, which is also moving in space. The effect of the earth's rota- 
tion on wind direction has the dimensions of acceleration. It is usu- 
ally expressed as (2VQsin<|>). From a basic law of physics, we recall 


that a force equals mass times acceleration (F = ma). The Coriolis 
acceleration becomes a force when applied to a mass of air. It was 
first expressed mathematically by a French scientist, G. G. Coriolis, 
in 1844. It can be shown that: 

C = 2VQsin$, 

where C is the Coriolis acceleration, V is the velocity of the wind, Q 
is a constant ( angular velocity of the earth's rotation ) and 4> is the 
latitude where the motion occurs. It can be recognized that the 
magnitude of C depends on the velocity of the wind and the lati- 
tudinal location. For a given wind velocity, C is zero at the equator 
and increases toward the poles, because the value of sin $ varies 
from to 1 from the equator to the poles. 

The Coriolis force acts at right angles to the horizontal direction 
of the wind. It does not, however, have any effect on the speed of 
the wind. 

The same effect is present in all motions relative to the surface 
of the earth, but it is inappreciable in most phenomena encountered 
in everyday experience because they are on a comparatively small 
scale. It does have to be allowed for, however, in calculating the 
motions of projectiles fired from long-range guns. It is of predomi- 
nant importance in considering the larger movements of the atmos- 
phere, as will appear in the two following chapters. Hence, the fact 
that moving air always tends to deviate to the right in the Northern 
Hemisphere and to the left in the Southern Hemisphere should be 
definitely fixed in mind. It should also be remembered that the speed 
of the wind is not affected by the deflecting influence. 

The three forces affecting moving air. Under a constant differ- 
ence of pressure, the pressure gradient tends to move air in a straight 
line, but as soon as motion begins, the effect of the earth's rotation 
is to cause it to move in a curved path. When the curving motion 
begins, a centrifugal force is developed, tending to pull the air out- 
ward from its center of curvature. Like the Coriolis force, the cen- 
trifugal force is not a true force in the physical sense. It is more 
properly called centrifugal action or reaction. Hence, the movement 
of the air is the resultant of three influences acting simultaneously, 
namely, the pressure gradient force, the earth's deflection, and the 
centrifugal force due to the curvature of the path with reference to 



the earth. These forces are illustrated in Fig. 72. The force p, repre- 
senting the pressure gradient, keeps a constant direction. The cen- 
trifugal and deflective forces, c and d, are always perpendicular to 
the instantaneous direction of the wind. The force c is opposite to d 
when the pressure gradient is from a center of high pressure out- 
ward, and in the same direction as d when the pressure gradient is 
inward to a center of low pressure. 










Fig. 72. Three Forces Affecting the Wind. Pressure gradient, ;>, starts the wind in 
motion and is balanced by the centrifugal force, c, and the Coriolis force, d, to create 
a gradient wind. 

Gradient winds. It can be shown, mathematically, that the re- 
sultant direction of motion, under these three forces, is along the 
isobars instead of across them, and the resultant speed, when a 
steady state is reached, is such that the centrifugal and deflective 
forces together balance the horizontal pressure gradient. A wind 
moving along the isobars at such a velocity that the force due to 
pressure gradient is balanced by deflective and centrifugal effects, 
is called a gradient wind. Its instantaneous direction at the point of 
the forces is represented by the line w, parallel to the isobars. The 
gradient wind results directly from the pressure gradient, since the 
other forces exist only after the gradient has initiated the air move- 
ment* A special case exists when the isobars are straight and par- 
allel and the gradient does not change over large areas. The cen- 
trifugal force is negligible and the gradient force is balanced by the 
Coriolis force alone. The wind is then called a geostrophic wind. 


The gradient wind can be calculated when the pressure gradient 
is known. At heights of 1,500 feet and more above the surface, the 
actual wind closely approximates the calculated gradient wind, both 
in direction and in speed. These winds flow perpendicularly to the 
pressure gradient (parallel to the isobars), directed toward the right 
of the pressure force in the Northern Hemisphere and toward the 
left in the Southern Hemisphere. 

Surface winds. Turbulence and friction near the earth's surface 
reduce the speed produced by a given pressure gradient. Friction 
has the effect of a new force acting in a direction opposite to the 
direction of the wind. With reduced wind speed, the deflective and 
centrifugal forces are less, whereas the pressure force remains the 
same. The resulting winds are, therefore, pulled around slightly in 
the direction of the pressure force, as indicated by Fig. 73. Wher- 
ever there is friction, the wind will tend to move across the isobars 
in the direction of the pressure gradient. 

j 1005 Mb 

-1002 Mb 

999 Mb 


Fig. 73. Surface Winds in Relation to the Isobars. Surface friction, /, disturbs the 
balance between the other forces and causes the wind to flow slightly across the 
isobars toward low pressure. When the isobars are straight, there is no centrifugal 

As a result of all these influences, we have the following general 
rule for the movement of the lower air: In a region of low pressure, 
the air has an inward-curving motion in a counterclockwise direc- 
tion in the Northern Hemisphere, and clockwise in the Southern 
Hemisphere. This is called a cyclonic circulation. From a region of 
high pressure, the air moves spirally outward in a clockwise direc- 
tion in the Northern Hemisphere, and in the opposite direction in 
the Southern Hemisphere. This is an anticyclonic circulation. In 
1857, Buys-Ballot gave the following practical rule for determining 
the distribution of pressure from the wind direction: If you stand 



with your back to the wind, pressure is lower on your left than on 
your right in the Northern Hemisphere, and the reverse in the 
Southern Hemisphere. This is known as Buys-Ballot's law. 

Winds Due to Local Temperature Differences 

The following well-known winds develop under special circum- 
stances and on a small scale as compared with the large movements 
of the air. They serve well to illustrate the direct relations among 
temperature, pressure, and air movement. Some of them are due to 
unequal heating of the air; in others, the initial impulse is given by 
the loss of heat, that is, by cooling. 

Sea breeze. Along the seacoasts, the land warms more than the 
adjacent water by day in summer sunshine. The warmed air over 
the land expands, bending the isobaric surfaces upward, and air 
flows out over the ocean from the upper surface of the expanded air 
(Fig. 74). This effect decreases the pressure over the land surface 
and increases it over the water, so that, when the circulation is es- 
tablished, the lower isobaric surfaces bend downward over the land 




Fig. 74. Sea Breeze. The first result of heating the land is the upward bending of 
the isobars over the land, producing an upper-level high; the second is the seaward 
flow of the upper air, reducing the surface pressure over the land and increasing it 
over the water; the third is the beginning of the inflow from the sea. 

and upward over the water, and the reverse in the upper part of 
the expanded layer. Thus is set up a movement of the lower air from 
the ocean to the land. This is the sea breeze, a partial convectional 
circulation. The circulation is incomplete because the air that flows 



out over the ocean from the top of the expanded layer spreads out 
broadly, and the downward movement is slow and distributed over 
a large area. As a result, little of the original air returns to the land, 
but the lower air along the surface of the ocean flows inland. 

Only a shallow layer of the air is affected by these changes; the 
sea breeze is usually not more than 800 to 1,200 feet (240-370 m) 
deep. It begins, usually about 10 A.M., some distance off shore and 
gradually extends inland to a distance of from 10 to 30 miles ( 16-48 
km), and seaward about the same distance. Toward evening it 
begins to subside. At places around the Great Lakes, notably on the 
western shore of Lake Michigan, there is, in summer, a similar lake 
breeze, which, however, extends inland only 2 or 3 miles ( 3-5 km ) . 

Sea breezes have an important moderating effect on the tempera- 
ture of coastal regions. Where the sea breeze is of daily occurrence, 
as in parts of California in summer, the afternoon temperatures 
average materially lower than they otherwise would, and the least 
agreeable part of the day is often in the forenoon before the breeze 
arrives. The cities of Chicago and Milwaukee have two summer cli- 
mates, one within a mile or two of the lake shore, and a consider- 
ably warmer one a few miles back from the lake where the lake 
breeze does not reach. 

Land breeze. At night the land cools more than the water, the 
air over the land becomes denser than that over the water, and the 
isobaric surfaces aloft slope downward toward the land (Fig. 75). 




Fig. 75. Land Breeze. The first result of cooling is the settling of the isobaric sur- 
faces over the land, producing an upper-level low; the second is the landward flow 
of the upper air, increasing the surface pressure over the land; the third is the be- 
ginning of the surface breeze from the land. 


When this occurs, air from the ocean begins to flow inland at the 
top of the cooled mass, thus increasing the pressure over the land 
and starting a movement out to sea at the surface. This is the land 
breeze, a wind due to local cooling. Again the circulation is incom- 
plete; the vertical movements are so diffuse and gentle as not to 
constitute perceptible currents. The land breeze is usually less de- 
veloped than the sea breeze; it is shallower, has less speed, and ex- 
tends only 5 or 6 miles ( 8-10 km ) over the sea. The principal rea- 
son for this effect is that temperature differences between land and 
water surfaces are less by night than by day. The effect of the land 
breeze is to remove the cooled air and to prevent the temperatures 
falling so low as they would if the air remained in place. In tem- 
perate latitudes, sea and land breezes are most frequent in summer 
and when skies are clear or have only scattered clouds. In tropical 
regions, they are frequent throughout the year. 

Valley breeze. The heating of a valley floor and its slopes by 
day sometimes results in a slow movement of warmed air up the 
valley or up the sides of mountains, The isobaric surfaces bend up- 
ward over the valley, and the air flows toward the sides. With spe- 
cially favorable topography to concentrate the movement, a strong 
up-valley breeze may develop by day. In some coastal valleys, sea 
and valley breezes combine to produce strong winds. 

Mountain breeze. Air on mountainsides and sloping plateaus 
cools by night more rapidly than the free air at some distance from 
the slopes or than the air in the valleys below. The draining of 
cooler, denser air down the slopes into the valleys, under the ac- 
tion of gravity, is somewhat analogous to the flowing of water down 
hill, but the air spreads out from the mountainsides, as water does 
not, and mixes with other air. The downward movement results in 
dynamic warming which, by lessening the density of the air, retards 
its flow. Hence, the movement is usually slow, but the warming 
effect of descent is more than offset by radiation cooling. The cold 
air may collect in pockets in the valleys and produce inversions of 
temperature, so that in the end (by morning) the valley bottoms 
are colder than the hillsides from which the cold air has been dis- 

The air may converge in narrow canyons and then gain consid- 
erable velocity and extend outward a few miles from the mouth of 
the valley. In these cases, radiation cooling, which proceeds slowly, 


may be largely counteracted by adiabatic warming. In Utah the 
effects of this warming and of the turbulent mixing of the air by 
reason of fairly rapid motion are sufficient to prevent early frosts in 
autumn and thus to prolong the growing season on the bench lands 
at the mouths of canyons. 1 In some mountain regions, the upslope 
valley breezes result in the formation of cumulus clouds and daily 
afternoon showers during the summer. The rain ceases and the skies 
clear toward evening, as the valley breeze weakens and the moun- 
tain breeze begins to develop. 

Katabatic winds. Along the northern coast of the Adriatic Sea, 
a plateau region rises at the rear of a narrow coastal plain. In the 
winter, the air over this plateau sometimes becomes quiet and cold 
by radiation cooling. It then flows down the slopes as a cold, north- 
east wind, known as the bora. The bora occurs by either day or 
night, but is most frequent and strongest in the latter part of the 
night. A similar cold wind, coming from the higher and often snow- 
covered land to the north, occurs during the winter on the Mediter- 
ranean coast of France, where it is called the mistral. The bora and 
mistral are fully developed only when the pressure gradient has a 
southward component. Such winds as the mountain or canyon 
breezes and the bora and mistral are given the general name of kata- 
batic winds, gravity winds, or fallwinds, due to the flowing of cold, 
dense air downslope under the pull of gravity. Fallwinds are com- 
mon along the Norwegian coast, and violent katabatic winds often 
descend from the glacier-covered interiors of Greenland and Ant- 

Sea breezes and valley breezes result from daytime heating; land 
and mountain breezes, from nighttime cooling; katabatic winds in 
general, from radiation cooling, whether diurnal or of longer period. 
Hence all are clear-weather phenomena, and all are rather shallow. 


Just as along the coastlines the relation between land and water 
temperatures changes daily under the influence of insolation by day 
and earth radiation at night, so in the longer period of a year there 
are seasonal temperature differences between entire continents and 

*W. B. Hales, "Canyon Winds of the Wasatch Mountains/' Bulletin American 
Meteorological Society, Vol. 14, Aug.-Sept, 1933; pp. 194-196. 


oceans. Continents are warmer than oceans in summer and colder 
in winter. The resulting tendency is to develop over continents rela- 
tively low pressure in summer and high pressure in winter. 

Seasonal temperature differences set up convectional circulations 
analogous to sea and land breezes but having an annual, instead of 
a diurnal, period. The wind tends to blow toward warm continental 
interiors in summer and from cold land areas in winter. These winds 
are called monsoons. Monsoons are winds that reverse their direc- 
tion with the seasons under the influence of seasonal temperature 
differences between continents and oceans. They are best developed 
in eastern and southern Asia, where larger phases of the movement 
of the air are also involved, as will be noted later. They are promi- 
nent in equatorial Africa and occur to some extent in Australia, the 
Spanish peninsula, and other places. In a large portion of the in- 
terior and eastern United States, the prevailing winds change from 
southerly in summer to northerly in winter. This reversal of the 
winds is a monsoon effect, and it is an important factor in the cli- 
mate of the central and eastern states. It results in the presence of 
much warm and humid air from the tropical Atlantic and from the 
Gulf of Mexico in summer, and of much cold, dry air from the in- 
terior of Canada in winter. 

Polar-Equatorial Air Movements 

Because of the great and permanent temperature contrast be- 
tween tropical and polar regions, we might expect to find a con- 
vectional circulation, analogous to a sea breeze or a monsoon, 
between equator and poles. On a uniform, nonrotating globe there 
probably would be a continuous circulation of this kind, with sur- 
face winds blowing toward the equator and upper winds toward 
the poles. On the rotating earth there is, indeed, an interchange of 
equatorial and polar air, but not a simple continuous exchange in 
a closed path. The existing temperature differences on the earth do 
compel movements of air between high and low latitudes, but a 
number of factors serve to make these movements complex and 

Modifying influences. The deflection due to the earth's rotation 
prevents a simple north-south interchange of air. Winds that start 
as south winds in equatorial regions become west winds in north- 


ern latitudes, and north winds from Arctic regions become east 
winds. The lack of uniformity in the change of temperature from 
equator to poles is another factor that prevents a simple interzonal 
transfer of air. Diversities of the earth's surface, such as the irregular 
distribution of land and water, the variations in elevation of the 
land, and differences of surface covering, all result in local or wide- 
spread temperature differences which prevent the development of 
a continuous temperature gradient between equator and poles, 
These temperature divergences create local pressure gradients not 
related to latitude, as is illustrated by the monsoons, sea and land 
breezes, and other local winds. In the third place, these tempera- 
ture relations change with the seasons, as the sun's rays shift 
north and south, thus interfering with a continuous circulation. 
Finally, there are irregular, moving disturbances of the atmosphere, 
to be discussed later, which render the actual zonal interchange of 
air still more complex. 

All the winds discussed in this chapter are examples of the propen- 
sity of surface air to blow toward a warm area or away from a cold 
area. They also serve to exemplify Humphreys' concisely stated 
general principle: "Atmospheric circulation is a gravitational phe- 
nomenon, induced and maintained by temperature differences/' 


1. Given the following sea-level pressure readings in millibars: Omaha, 
1029; Des Monies, 1023; Davenport, 1015; Chicago, 1004; and the fol- 
lowing distances between cities: Omaha to Des Moines, 120 miles; Des 
Moines to Davenport, 140 miles; Davenport to Chicago, 130 miles. 

(a) What is the pressure gradient per 100 miles between each of 
the adjacent cities? 

(b) If the wind velocity hetween Omaha and Des Moines is 15 
miles per hour, what is the velocity between the other cities, 
assuming it to be proportional to the gradient? 

(c) From what general direction i$ the wind blowing? 

2. Draw diagrams illustrating the pressure changes and air movements 
occurring in sea, mountain, and valley breezes. 

3. Small cumulus clouds often occur with a sea breeze but not with a 
land breeze. Why? 

4. Look up the meaning of the sine of an angle and explain why the 
Coriolis force is zero at the equator. 


5. (a) Draw two sets of two concentric circles. Label one set "high 
pressure" and the other "low pressure." Now draw arrows to or from the 
four points of the compass to indicate the direction of the pressure force 
in each case. Dash in a deflection to the right of each arrow. This should 
show the direction of wind flow around high- and low-pressure areas in 
the Northern Hemisphere. 

(b) Repeat this operation for the Southern Hemisphere. 




Pressure and winds are different phases of the same large problem 
dealing with the distribution of the air over the earth, its changes 
in distribution, and the processes by which the transportation of 
great masses of air is achieved. This constitutes the central problem 
of meteorology, concerning which many details remain unknown 
because of the great extent of the atmosphere and the variety and 
complexity of the influences affecting its movements. 

Observations show that there are large areas of the earth where 
the winds are predominantly from one direction throughout the 
year, other areas where the prevailing direction changes with the 
seasons, and still others where the winds are so variable from day to 
day that no systematic movement is evident to the ordinary ob- 
server. Related to the variability of the wind direction is the fact, 
previously noted, that pressures are also changeable. Hence, it might 
be inferred that no simple, permanent plan of distribution of pres- 
sure and wind exists. Nevertheless, if we take the average annual 
pressure and the prevailing winds, over the globe, we find not only 
that pressure and winds are closely related, but also that their dis- 
tribution may be generalized into a simple system, dividing the 
earth into a few large zones or belts. The average general distribu- 
tion of wind movement is known as the general circulation. 

Yearly Averages of Pressure 

The mean annual pressures over the globe are represented in 
Fig. 76. For polar regions, both in the Arctic and the Antarctic, data 
are meager and the isobars are doubtful. Especially over Antarctica, 




a e 







pressure is uncertain, but it seems probable that it is considerably 
higher at the center of the continent than along the edges, where 
a pressure of 29.3 inches (992 mb) is indicated. A study of the chart 
will disclose the following alternating zones of high and low pressure. 

Equatorial belt of low pressure. In equatorial regions there is a 
belt where the pressure is less than 29.9 inches (1,013 mb) through- 
out, and less than 29.8 inches (1,009 mb) in parts of the Eastern 
Hemisphere. The belt varies in width, but completely encircles the 
earth. On the average, its center is somewhat north of the equator. 
Within the equatorial belt, the winds are generally light and vari- 
able, with frequent calms, but with an average slow drift from east 
to west. The entire belt is called the doldrums, but this word ap- 
plied originally only to the ocean areas near the equator, where 
sailing ships were frequently becalmed. 

Subtropical high-pressure belts. Centered at about 35 north 
and 30 south latitude, there are irregular belts where the average 
pressure is above 30 inches (1,016 mb) and within which are cer- 
tain areas averaging more than 30.1 inches (1,019 mb). These are 
the subtropical high-pressure belts or the horse latitudes. 1 

The name subtropical high is applied especially to the centers of 
higher pressure within the belts. The northern belt, where large 
land and water surfaces alternate, is more irregular than the south- 
ern belt, which is largely over water and therefore under a more 
nearly uniform influence. These belts are regions of variable winds, 
averaging light and changing with the seasons. They are sometimes 
invaded by traveling disturbances attended by stormy winds. 

The equatorial belt of low pressure and subtropical belts of high 
pressure may be explained as the expression of a convectional cir- 
culation, air rising in the heated doldrums, moving poleward aloft 
in both hemispheres, being deflected eastward by the earth's rota- 
tion, and finally accumulating and settling in the belts of higher 
pressure, out of which winds blow toward the equator (Fig. 77), 
The movements are probably not so simple and direct as this ex- 
planation implies, but undoubtedly convectional movements some- 

1 The name horse latitudes is said to have arisen in the days of sailing ships, when 
several Spanish vessels were becalmed at the center of the subtropical high-pressure 
circulation. The cargoes, consisting mostly of horses destined for the New World, 
were dumped overboard to lighten the ships and increase their mobility. 



thing like this do occur. These belts are a prominent and permanent 
feature of the general circulation. 

Direction of flotation 










Fig. 77. Pressure and Wind Belts of the World. Drawing hy Ed Raids. 

Polar low pressure. There is a continuous belt of low pressure 
in the Southern Hemisphere between latitudes 60 and 70. This 
belt overlies a water surface. In corresponding latitudes in the 
Northern Hemisphere there are large, cold land masses, and their 
effect is to increase the pressure; but over the northern oceans there 
are well-defined areas of low pressure. These are centered in the 
vicinity of the Aleutian Islands in the Pacific and between Green- 
land and Iceland in the Atlantic. Winds from the west or southwest 
blow into these regions of low pressure from the equatorward side 
in accordance with the pressure gradient as modified by the deflect- 
ing influences. 

Polar caps of high pressure* In the center of Antarctica there 
appears to be a permanent cap of high pressure, but data are still 


inadequate to give an exact figure for the annual mean. All explorers 
of Antarctica have reported frequent, strong, southeast winds from 
the interior, and these reports confirm the existence of relatively 
high pressure near the pole. In the Northern Hemisphere, the cap 
of high pressure probably is not centered at the pole but extends 
from northern Greenland westward across the northern islands of 
Canada. Here, too, data are meager. Easterly winds blow out of 
these caps of high pressure. 

Circulation zones and cells. In conformity with this general dis- 
tribution of the pressure in alternating belts of high and low pres- 
sure, the general circulation is divided into three zones in each 
hemisphere. One of these is the zone between the subtropical high- 
pressure belt and the equator, in which winds move equatorward 
with a large component from the east. The second zone lies between 
the subtropical high-pressure and the polar-circle low-pressure belts, 
that is, between latitudes 30 and 60, approximately, in each hemi- 
sphere. In this zone, the air moves poleward but by deflection be- 
comes largely westerly. Finally, in the third zone, the air moves 
out of the polar cap of high pressure toward the lower pressures at 
about latitude 60, becoming easterly by deflection. Thus, instead 
of a continuous circulation between equator and poles, such as 
would occur on a uniform, nonrotating earth, we find each hemi- 
sphere divided into three more or less independent circulation zones 
(Fig. 78). 



90* SOT 3or o* 3cr ecf 90* 


Fig. 78. Schematic Representation of the- Cells of Atmospheric Circulation. Drawing 

by F. ]. Williams. 

A closer examination of the pressure belts shows that they are not 
of uniform pressure throughout, but that they are divided into a 
number of centers, or cells. There are centers of low pressure near 
the equator and near latitude 60, and centers of high pressure in 
the subtropical belts of both hemispheres and near the poles. These 


centers of high and low pressure control, or at least influence, the 
air movement around them, resulting in the formation of several 
cellular circulations, some cyclonic, others anticyclonic, superim- 
posed upon the zonal circulations. The result is a meridional (north- 
and-south) movement of the air on the east and west sides of these 
cells. But on the whole, as a result of all the complex influences, the 
latitudinal ( east-and-west ) circulation is much greater. 

The cells just mentioned are known as centers of action because 
it is along their boundaries that most storms originate arid travel. 
They are also called semipermanent centers of high and low pres- 
sure because they tend to persist in the same general regions, but 
their exact position and intensity may change somewhat with time. 
The entire zones shift northward in the northern summer and south- 
ward when it is summer in the Southern Hemisphere. They follow 
the sun and the seasonal changes, but they lag a month or two be- 
hind the sun. The seasonal change in the position and intensity of 
the cells is a monsoonal effect. The cells of low pressure tend to 
migrate in summer to the heated continental interiors from the 
cooler oceans, and the pressure is relatively higher over the oceans. 
In winter, the cells of high pressure are centered over the cold con- 
tinental areas and low pressure is intensified over the relatively warm 
ocean waters. 

January and July Averages of Pressure and Winds 

The major differences in pressure distribution and wind direction 
between winter and summer are shown in Figs. 79 and 80. Let us 
now examine these charts and note in some detail the outstanding 
features of the general circulation and its component parts, the zonal 
and cellular circulations, as these are modified by the distribution 
of land and water and by the seasonal variations in insolation. 

Doldrums. In January, the continuous equatorial belt of low 
pressure has its centers of lowest pressure over the land areas in the 
Southern Hemisphere, where it is midsummer. Note the centers in 
equatorial South America, equatorial Africa, and northern Australia 
(Fig. 79). In July, the belt is almost entirely north of the equator, 
and low pressure extends far northward over North America and 
Asia, with minima in northwestern India and southwestern United 
States (Fig. 80). Within the doldrums, the air movement in the 




lower atmosphere is mostly from an easterly direction, but note that 
there is a shifting between northeast and southeast with the sea- 
sons, as the center of the low pressure moves south and north, In 
January, northeast winds of the Northern Hemisphere extend to, 
and in some cases south of, the equator. In July, winds from the 
Southern Hemisphere cross the equator and reach 10 to 20 north 
latitude. The convergence of these winds in the doldrum region and 
the resulting vertical movements cause frequent and heavy rains 
throughout the year. 

High-pressure belts. In January, the subtropical high-pressure 
belt is practically continuous in the Northern Hemisphere near lati- 
tude 30, with somewhat higher pressure in the eastern parts of the 
Atlantic and Pacific than in the western parts of these oceans. In 
the Southern Hemisphere, where the land is warm in January, there 
are three maxima over the relatively cool oceans, in each case where 
the ocean water is abnormally cold for the latitude because of cold 
ocean currents moving northward. 

In July, in the Northern Hemisphere, the high-pressure belt is 
broken by the development of low pressure over the hot interior 
regions of southwestern United States and southwestern Asia, but 
there are well-developed and extensive cells of high pressure over 
the cool ocean areas. The cell in the eastern Pacific is known as the 
Pacific high, or Pacific anticyclone, and that in the eastern Atlantic 
as the Azores high, or Azores anticyclone. These two cells are of 
great importance in their influence on the weather of all temperate 
regions of the Northern Hemisphere. South of the equator, although 
pressure has risen over the land areas, the centers of highest pres- 
sure remain over the oceans, as in January. The small proportion 
of land in these latitudes is not sufficient to reverse the pressure dis- 
tribution as in the northern half of the world. 

Trade winds. Between the doldrums and the belts of higher 
pressure, there are steady, moderate winds, known as trade winds, 
blowing out of the high-pressure areas toward the equator. As they 
move equatorward, they are deflected to the west and become 
northeast trades in the Northern Hemisphere and southeast trades 
in the Southern Hemisphere. They are best developed on the eastern 
sides of the oceans, in air flowing out of the oceanic peaks of high 
pressure. In these situations, they are remarkably constant in direc- 
tion and speed, blowing almost uninterruptedly, day and night, win- 


ter and summer, with a velocity of from 10 to 15 miles per hour (5-8 
meters per second). Such steady winds do not occur over large land 
areas, and even in the western portions of the oceans, the trades are 
less constant. They are confined to the belt between 30 north and 
30 south latitude. 

Aleutian and Iceland lows. In contrast to the continuous belt of 
low pressure near the Antarctic circle, there are two distinct cells 
of low pressure near the Arctic circle. That in the north Pacific is 
known as the Aleutian low, and that in the north Atlantic, the Ice- 
land low. Both of these are strongly developed in winter, for each 
is in a region where the temperature of the water is raised by warm 
ocean currents, and each is near large land masses that become very 
cold. In summer, Alaska and Siberia become decidedly warmer 
than the adjacent waters, thus reversing the temperature gradient. 
This is followed by a reversal of the pressure gradient; the center of 
low pressure moves to the continents, and the Aleutian low prac- 
tically disappears. In the Atlantic area, Greenland, Iceland, and 
northwestern Europe remain comparatively cool during the sum- 
mer; they do not warm sufficiently to destroy the Iceland low. The 
Aleutian and Iceland lows exercise an important influence on the 
weather of North America and of Europe, respectively, as will be 
noted later. 

Continental highs of winter. As indicated in the preceding para- 
graph and as shown in Fig. 79, the Aleutian and Iceland lows are 
separated in winter by areas of high pressure over the continental 
interiors. The regions of highest pressure are Mongolia in the cen- 
ter of Asia, and the Mackenzie Valley in northwestern Canada. 
These are regions of intense winter cold, and the high pressure is 
largely a response to monsoonal influences. In other words, the con- 
trast between the abnormally warm waters of the Bay of Alaska 
and the north Atlantic, on the one hand, and the cold continents, 
on the other hand, results in the development of the strongly con- 
trasting areas of high and low pressure in these latitudes. 

Prevailing westerlies. Winds blowing out of the poleward sides 
of the subtropical belts of high pressure are deflected as they move 
into higher latitudes and become southwest winds in middle north- 
ern latitudes, and northwest or west winds in middle southern lati- 
tudes. These are known as the prevailing westerlies. They begin 


about 35 north and south latitudes and extend to the subpolar lows, 
in the vicinity of the polar circles. Near the surface of the earth, 
they are subject to many interruptions by storms and irregular, in- 
termittent winds from all directions, but the prevailing direction is 
from the west. They are often called the stormy westerlies. At the 
cirrus-cloud level, they come more steadily from a westerly direc- 
tion. These winds persist throughout the year but are stronger in 
winter, especially in the north Atlantic and north Pacific, where the 
deepening of the Aleutian and Iceland lows and the building up of 
high pressure over the continental interiors create steep pressure 
gradients. The area between latitudes 40 south and 50 south is 
almost entirely water, and the prevailing westerlies are strong and 
persistent throughout the year. The region is called by sailors the 
"roaring forties." 

Polar easterlies. Winds blowing out of the Antarctic cap of high 
pressure and deflected to the left are known as polar easterlies, 
While there are no winds blowing regularly from the sea around 
the north pole, there are prevailing out-flowing easterly winds from 
Greenland and, in winter, from the cold centers of Siberia and 
Canada also, and these may be considered as representing the polar 
easterlies of the Northern Hemisphere. Recent observations in north- 
ern Alaska show that the prevailing winds are from the east below 
3,000 meters ( about 2 mi. ) , and from the west above that height. 

Polar front. The relatively warm prevailing westerlies meet the 
cold polar easterlies or the cold air from continental interiors along 
an irregular shifting boundary which is known as the polar front. 
The polar front is the boundary surface of the cold air as it advances 
toward warmer latitudes. From day to day, this boundary changes 
its position, swerving far northward or southward, especially in the 
Northern Hemisphere in winter. At times warm air swings north in 
the Atlantic to northern Scandinavia and Spitzbergen; at. times cold 
air streams southward from northern Canada or northern Eurasia, 
chilling the Gulf coast of the United States or the Mediterranean 
coast of Europe. Perhaps some of this air from polar regions finally 
enters the trade winds and reaches the equator. It is along this mov- 
ing polar front that the storms, or barometric depressions, charac- 
teristic of the weather outside the tropics, develop. 


Winds Aloft 

Our direct knowledge of the pressure and movements of the upper 
air comes from observations of cloud movements, especially of the 
cirriform clouds, from pilot-balloon observations, and from the rec- 
ords of instruments carried aloft. Additional data are accumulating 
from these sources, but as yet our knowledge of the normal circula- 
tion of the upper air remains incomplete. To get a picture of the 
distribution of pressure and winds at different levels, it is possible 
to supplement the observations with calculations based on assumed 
lapse rates. The increasing number of radiosonde and rawin obser- 
vations is making these calculations unnecessary, especially over the 
United States. 

Upper troposphere. At upper levels in the atmosphere, the pres- 
sure pattern is much simpler than in the surface layers. The migra- 
tory highs and lows at the surface are usually shallow and lose their 
identity within the first mile or two of altitude. The resulting pres- 
sure pattern becomes latitudinal as the isobars trend generally east- 
west in response to the persistent belts of high and low pressures. 
It may be necessary to point out that winds normally increase in 
velocity with height above the earth's surface. This is generally true, 
at least up to the tropopause. 

With practically no friction aloft, the winds respond to the exist- 
ing forces to create gradient winds or geostrophic winds, accord- 
ing to the pressure pattern. Although there is considerable zonal 
transport of air at high levels as the winds meander over pressure 
troughs and ridges, the principal air movement is west-to-east 
(Fig. 81). 


Fig. 81. East- West Transport of Air in the General Circulation. 

Antitrades. The height to which the trade winds extend varies 
in different parts of the world and at the same place at different 
times of the year. The observed heights range between 3,000 and 


13,000 feet (1-4 km). Above the trades, a direct reversal of the 
wind direction has sometimes been observed. These winds, from 
the southwest over the northeast trades and from the northwest 
over the southeast trades, are known as the antitrades. It was for- 
merly believed that the antitrades represented a simple pattern of 
wind flow as a direct counterpart to the surface trades. Recent in- 
vestigations in the tropics have revealed, however, that the upper- 
air circulation in the tropics is much more complex than was once 
supposed. 2 The antitrades are not so constant as the trades, nor are 
they always present over them. 

As the antitrades reach latitudes 30 north and south, they have 
been cooled by radiation and expansion and they have become de- 
flected by the earth's rotation into a westerly current. This flow of 
cold air aloft tends to build up the subtropical high-pressure belts 
and encourage subsidence in these latitudes. 

Jet stream, A recently discovered atmospheric phenomenon, and 
one that may prove quite valuable in weather forecasting, is the jet 
stream. It is a narrow, meandering band of swift westerly winds that 
circles the globe in middle latitudes. Evidence of its existence was 
noted in 1922 when a weather balloon, released in England, came 
down only four hours later near Leipzig, Germany, some 570 miles 
(900km) away. 3 

During World War II, pilots flying bombing missions from Pacific 
bases to Japan encountered very strong head winds that almost 
arrested their flight and prevented them from reaching the desig- 
nated targets. Subsequent research revealed the general character- 
istics of the jet stream. 

The jet stream is a core of extremely swift winds in the prevailing 
westerleis of middle latitudes. It may range from 25 to 100 miles 
(40-160 km) in width and up to a mile or two (2-3 km) in depth. 
Wind speeds of 300 miles per hour have been recorded. The jet 
stream follows a serpentine path, but the movements are not entirely 
erratic and seem to follow a definite cycle. The velocity is strongest 
where north-south temperature contrasts are greatest. This indi- 
cates that the jet stream is located along the polar-front boundary 
and is strongest on the east side of continents during the winter 

2 Herbert Riehl, Tropical Meteorology. New York: McGraw-Hill Book Co., 1954, 
pp. 23-24. 

3 D. H. Johnson, Weather (London), September, 1953, pp. 270-274. 


months. It is believed to have much influence in steering air masses, 
which in turn are responsible for the general surface weather con- 
ditions. 4 Because of turbulence along the fringes of the jet stream, 
it can be dangerous to pilots unfamiliar with it, but it is being suc- 
cessfully utilized for air travel. 

Latitudinal Interchange of Air 

Mixing processes. It is evident from the foregoing discussion of 
the general circulation that the air is not completely shut off into 
separate compartments, but that much transposition and inter- 
change of air occur from one pressure belt to another. Some of the 
processes by which this mixing is accomplished are: (1) by the 
shifting of the center of the doldrums north and south of the equa- 
tor, thus transferring air from one hemisphere to the other; (2) by 
the circulation of the air around the semipermanent centers of high 
and low pressure as they gradually change their positions with the 
changing seasons. For example, the centers of high pressure in the 
Northern Hemisphere change from the continents in winter to the 
oceans in summer, involving a change in the direction and the des- 
tination of large masses of air; (3) by the movement of great quan- 
tities of cold air equatorward and of warm air poleward along the 
polar front as it rapidly alters its position; (4) by the rising and set- 
tling of air in various parts of the world, displacing surface air with 
air from aloft and involving both in new circulations. 

By such processes the air is kept mixed to a considerable extent, 
and this interchange of air has an important equalizing effect on the 
temperature. It is true that masses of warm air do accumulate in 
equatorial regions and cold air in polar regions, and the mixing is 
not perfect. If it were not for the winds, aided by the ocean cur- 
rents, the equatorial zone would become unbearably hot, for it re- 
ceives more heat than it radiates, and the temperature of the polar 
regions would fall extremely low in the long polar nights. 

The Asiatic monsoon. In the coastal region of China and in 
southeastern Asia and India, there is a complete reversal of wind 
directions with the seasons. This Asiatic monsoon furnishes a strik- 
ing example of the transfer of air across the equator with the migra- 

4 Jerome Namias, "Pattern of Wind Flow in the Upper Westerlies," Scientific 
American, Vol. 187 (June, 1951), pp. 26-31. 


tion of the sun. See the charts of January and July pressures and 
winds (Figs. 79 and 80). In the winter months, pressure is high over 
the cold interior of Asia, with inflowing air aloft. At the same time, 
the doldrums are south of the equator, with lowest pressure over 
northwestern Australia and adjacent islands. During these months 
the lower air, therefore, moves outward from the interior in an anti- 
cyclonic circulation and crosses the equator into the Southern Hemi- 
sphere. The wind is from the northwest in northern China, gradually 
changing to north in southern China, and then to northeast in Indo- 
China, India, and the northern Indian Ocean. Finally, it becomes 
northwest again after it crosses the equator and becomes subject to 
a Coriolis force directed to the left. Where the air does not pass 
over extensive water surfaces, it is very dry, because it has not been 
exposed to moisture sources. 

This is the northeast ( winter ) monsoon of India and Indo-China. 
The coldest air is shut off by the mountains, and the air that reaches 
this region is warmed and further dried as it moves downslope from 
the northeast. In this season, therefore, the Indian Peninsula has 
moderate temperatures, light winds, and very little rain. In the lati- 
tude of Japan and eastern China, the winter monsoon is of moderate 
strength and subject to interruptions by traveling storms. 

In the summer, there is a continuous pressure gradient from the 
high-pressure belt in the southern Indian Ocean to the low-pressure 
area in heated southwestern Asia. The southeast trades cross the 
equator and become the southwest monsoons of the northern In- 
dian Ocean and of India and Burma ( Fig, 80 ). As the warm, moist, 
conditionally unstable air from equatorial waters moves northward 
toward the Himalayas, it is forced upslope, and heavy to excessive 
rains result. This is the hot, rainy, summer monsoon upon which 
the crops depend. 

Because of the steady, moderate breeze, which is stronger than 
the winter monsoon in this India-Burma region, the heat is less op- 
pressive than it would otherwise be. Between seasons, in spring and 
again in autumn, while the pressure distribution is changing, winds 
are light and variable, and also hot and humid. At these seasons 
living conditions are rather uncomfortable. North of the mountains 
there is practically no rain, even in summer. The summer monsoons 
are from the south in southeastern China and from the southeast in 
Japan and northeastern China, forming a cyclonic circulation around 


the interior area of low pressure. They are attended by moderate to 
heavy, but usually not excessive, rainfall. 

The movement across the equator. Not only in the Asiatic mon- 
soon, but also in the trade winds in the Atlantic and Pacific Oceans, 
there is movement of air across the equator, northward in the north- 
ern summer and southward in the southern summer. This move- 
ment of surface air to the warm hemisphere must mean that the 
pressure is higher in the cold hemisphere, and hence that there is 
more air over the winter half of the globe than over the summer 
half. The average pressures obtained by observation show this to be 
true. While surface air travels to the warmer half of the earth, it is 
more than replaced by air moving in at higher elevations and set- 
tling over the cold continents. 

Sources of energy. To move the masses of air involved in the 
general circulation requires an immense amount of work. Air, like 
all other material substances, has inertia; force is required to start 
it moving. After motion begins, friction and turbulence oppose the 
motion, reducing the speed of the wind and tending to break down 
the circulation and to make the air flow across the isobars into the 
areas of low pressure. Despite these opposing forces, such great cur- 
rents of air as the prevailing westerlies and the trade winds con- 
tinue without interruption. 

The following four factors contribute toward the accomplishment 
of this work: 

1. Insolation. The energy received from the sun results in an 
unequal warming of the air, largely because of an unequal warm- 
ing of surfaces with which the air comes in contact. This inequality 
is due in the first place to the differing amounts of insolation re- 
ceived, but the character of a surface also influences its temperature. 

2. Gravitation. The unequal heating of the air produces differ- 
ences in its density, and the force of gravity then causes the heavier 
air to seek the lower level, displacing the lighter air. 

3. Condensation. The latent heat released by the condensation 
of water vapor supplies much energy and is often responsible for 
vigorous upward convection. 

4. Rotation. The rotation of the earth results in changing the 
direction of the moving air and is responsible for the great amount 
of eastward and westward movement found in the general circula- 
tion. These are the things which produce the movements of the air, 


and which, as limited by inertia and friction, result in maintaining 
the general circulation as we find it. 

The primary source of energy is the sun, which causes convective 
movements of the air. Sir Napier Shaw expresses it thus: "There is 
nothing but thermal convection to act as the motive power for every 
drop of rain that ever fell and for every wind that ever filled a sail 
or wrecked a ship since the world began." 5 


1. Explain the occurrence of southwest winds near the surface over 
the Bay of Bengal in summer. 

2. Describe the characteristic weather of the doldrums; of the trade 
winds; of the subtropical highs; of the prevailing westerlies. 

3. Determine the latitude and longitude of the centers of low pressure 
in January and in July. 

4. Determine the latitude and longitude of the peaks of high pressure 
in January and in July. 

5. What is the difference in the weight of air over the United States 
in winter and in summer, assuming that the average pressure is 30.1 
inches in winter and 29.9 inches in summer? (A cubic inch of mercury 
weighs 0.49 pound. The area of the United States is approximately 
3,026,000 square miles. ) 

r> Sir Napier Shaw, The Air and Its Ways. London: Cambridge University Press, 
1933, p. 99. 



As noted in the previous chapter, there is a general circulation 
pattern which tends to persist from season to season over the earth. 
This pattern obviously does not remain fixed in all its details, even 
for a short period of days, or the weather would tend toward mo- 
notony in all parts of the world. 

Variations of intensity in the high- and low-pressure belts permit 
the accumulation of large masses of air over certain geographical 
regions. As a mass of air lingers over a single region without being 
replaced by new air, it tends to assume the temperature and humid- 
ity characteristics of that region. Conflict results when two such air 
masses finally move together from different source regions. Some of 
the principal weather characteristics of the middle latitudes are cre- 
ated in this manner. Air masses take a leading role in the weather 
drama. Present-day weather analysis and forecasting consist, to a 
large degree, of studying the structure and characteristics of air 
masses and their interactions when they converge. 

Air Masses 

Nature of air masses. An air mass may be defined as a large body 
of air of considerable depth which is approximately homogenous 
horizontally. At the same level it has nearly uniform physical prop- 
erties, especially as regards its temperature and its moisture con- 
tent. Such masses are formed over large uniform areas of land or 
water surface where the wind movement is light. Under these con- 



ditions, the air near the surface gradually takes on uniform charac- 
teristics, approaching those of the surface over which it lies, and 
the air above adjusts itself to the temperature and moisture condi- 
tions at the surface. The principal processes bringing about this ad- 
justment are radiation to and from the air, vertical convection, turbu- 
lence, and horizontal movement (advection). 

The warm waters of the Gulf of Mexico and the Caribbean Sea 
and similar areas in the Pacific Ocean between Mexico and Hawaii 
are areas over which great masses of warm air accumulate. These 
are regions of light winds on the edge of the trade-wind belt. The 
snow- and ice-covered area comprising northern North America and 
the adjacent portions of the Arctic Ocean is a source of extremely 
cold air masses .? Observations show that the movement of air in 
northern Alaska is from 30 per cent to 40 per cent less than in the 
United States. The same is probably true for the Mackenzie River 
Valley of northwest Canada. This region is therefore favorable for 
the accumulation of masses of cold air. 

Eventually the air masses are carried in the general circulation 
from their source regions to other parts of the world. Thus, warm, 
moist, tropical air is transported northward, and cold, dry, polar air 
southward. As they move, they tend to retain their properties, espe- 
cially in their upper portions. The surface layers are more or less 
modified by the surfaces over which they move. After the two 
masses from different sources meet, they tend to preserve their iden- 
tities, instead of mixing freely, and thus create "fronts" or "discon- 
tinuities" along the boundary zone. As a front crosses a given place 
on the earth, there is a more or less abrupt, discontinuous change 
in the properties of the air due to one air mass replacing another. 
It is along these fronts that the principal changes in weather occur. 
Of primary importance in its effect on the weather is the distribu- 
tion of temperature and moisture in the air masses. 

Classification of air masses. There is some lack of uniformity in 
the classification of air masses, but the following conforms to pres- 
ent American usage and includes the air masses that affect the 
weather of the United States. With reference to latitude of origin, 
air masses are divided into four types, namely: arctic (A), polar 
(P), tropical (T), and equatorial (E). The differences between 
arctic and polar air and between equatorial and tropical air are 
small and of little significance. In addition, it is seldom that either 



arctic or equatorial air affects the weather of the United States. For 
these reasons, they will be omitted from further discussion. 

Air-mass types are subdivided with reference to the nature of the 
surface over which they originate into continental (c) if the air 
mass originates over land, and maritime (m) if it has its origin over 
water. An air mass is further classified according to its low-level tem- 
perature relative to the surface over which it is passing. It is im- 
portant to the meteorologist to know whether the air mass is being 
heated or cooled from the surface level. Cooling from below favors 
stability, and heating favors instability. From surface observations, 
then, the air mass can be classified as warm ( w) or cold (k), mean- 
ing, respectively, that it is warmer or colder than the surface with 
which it is in contact. One additional type of air mass is generally 
recognized. It is very dry air descending from aloft and is known 
as a superior ( S ) or subsidence air mass. 




Source regions 

Polar continental 


Canada, Alaska, Arctic region. 

Polar maritime 


Northwestern Atlantic, North Pacific, 
particularly in vicinity of Aleutian 

Tropical continental 


Southwestern United States and north- 
ern Mexico, in summer only. 

Tropical maritime 


Sargasso Sea, Caribbean Sea, Gulf of 
Mexico, subtropical North Pacific. 



Upper levels of troposphere in region 
of subtropical highs. 

Modifications of air masses. When any one of these air masses 
moves from the area in which it acquired its characteristic proper- 
ties to a region of different surface conditions, it immediately begins 
to be modified by the new influences to which it is subjected; and 
the longer it remains under the new conditions, the more the orig- 
inal influences are neutralized. The lower layers, especially, undergo 
a gradual transition in temperature and humidity, while the upper 
layers remain more or less unchanged, unless the mass becomes 


unstable. The most significant changes in their effect upon the 
weather are changes in stability, resulting from changes in tempera- 
ture and in moisture content. 

It will be seen that the properties of an air mass depend upon its 
history. We need to know, first, the fundamental properties of the 
mass, as acquired at its source. Then we should know by what path 
it has reached its present position and the nature of the surfaces 
over which it has moved, and, finally, how long it has been away 
from its source and subjected to modifying influences. Air may reach 
the Ohio valley, for example, as a southwest wind which a few 
days earlier moved southward from Canada as true polar air. The 
weather that it brings will be quite different from that brought by 
a southwest wind originating as tropical Gulf air. The direction of 
the wind is not always a true indication of the history, nor, there- 
fore, of the properties of the air. 

One of the most frequent and most important modifications of air 
masses is that indicated by the addition of 'the letters w and k to 
air-mass designations. For example, when a cold, stable mP air mass 
moves southward over the heated interior of the United States in 
summer, it is much colder than the surface over which it is moving, 
and is designated mPk. It is thus heated from below, rather rapidly 
near the surface and more slowly and to a lesser degree aloft. This 
heating steepens the lapse rate, decreases the stability, and favors 
turbulence and convection. Thus the original character of the air 
mass and the kind of weather attending it are considerably modi- 
fied. Similarly, a mass of warm, tropical air moving northward over 
land in winter is cooled at its surface and made more stable, It is 
obvious, then, that heating and cooling of an air mass from below 
are important modifying influences. In general, w indicates a stable 
air mass and k an unstable air mass. 

Vertical movements are also important modifiers of air masses. 
Downward movement of stable air increases both the temperature 
and the stability, whether the descent is by movement downslope, 
or by direct subsidence from aloft. Lifting of an air mass results in 
adiabatic cooling, an increased lapse rate, and active convection in 
case there is convective instability to begin with. Addition and re- 
moval of water vapor are other processes by which air masses are 
modified. Water may be added by evaporation from a moist sur- 


face or from falling rain. Removal of water by condensation and 
precipitation adds latent heat, decreases lapse rate, and makes fu- 
ture condensation less probable. Mixing, either by turbulence or 
by the convergence of air currents, is another process by which the 
properties of air masses are modified. Thus, in studying the charac- 
teristics of air masses, it is to be remembered that they are subject 
to constant change as they move from their source regions. Hence, 
it is important to know their path and history. 

Characteristics of North American Air Masses 

The weather that we experience from day to day depends pri- 
marily upon the characteristics of the air masses that move over us. 
By "characteristics" we mean specifically the temperature, the lapse 
rate, and the moisture content of the air masses. The pilot is espe- 
cially interested, also, in such properties as dew point, visibility, 
ceiling, and kind of clouds. The characteristics are determined by 
the temperature and moisture conditions of the source regions over 
which the air masses were formed and by their subsequent histories. 
We shall now discuss briefly the characteristics of the principal 
North American air masses and how they differ from winter to 

Polar continental (cP) air masses in winter. The source regions 
for cP air in winter are Canada, the ice-covered Arctic Ocean, and 
northeastern Siberia. Such of these air masses as affect the United 
States usually originate in the high-pressure center over northwest- 
ern Canada. For a considerable period, this air has overlain a frozen 
or snow-covered surface and has become very cold in its lower levels 
by radiation cooling in the long winter nights of high latitudes. 
Since cooling in the free air is not so rapid as at the surface, and 
also since the air is, in general, subsiding, the temperature usually 
increases from the ground up to a considerable elevation. This in- 
version of temperature means marked stability; convection is im- 
possible and turbulence is reduced. In such cold air the moisture 
content (specific humidity) is necessarily very low, but relative hu- 
midity may be high. Generally, cP air brings clear, cold weather 
and good visibility to the United States. As these air masses move 
southward, they usually move over warmer surfaces and become cPk 
air masses. The consequent warming of the lower air is sometimes 


sufficient to cause instability, resulting in convective movements at- 
tended by cloudiness and precipitation. 

Polar continental (cP) air masses in summer. The source regions 
of summer cP air are in Alaska and in central and northern Canada. 
The ground is not snow-covered, and there is some heating of the 
surface in the long hours of summer sunshine, sometimes resulting 
in conditional instability. But the air usually remains cool as com- 
pared with surface temperatures farther south. The moisture con- 
tent is small and the relative humidity usually not over 45 per cent. 
Hence, such air reaches the United States with temperature and 
humidity both moderately low. It becomes cPk air and may become 
unstable, but the condensation level is high and the air usually re- 
mains cloudless. 

Polar maritime (mP) air masses in winter. There are two source 
regions of the mP air masses that occur in the United States in win- 
ter. One of these is the north Pacific Ocean in the region of the 
Aleutian low, and the other is the cold northwestern Atlantic off 
the coasts of Newfoundland, Labrador, and Greenland. In the main, 
the Pacific air masses were formed originally as stable cP air in 
Siberia; but as they move eastward over the relatively warm water, 
from which there is active evaporation, they become warm and 
humid in their lower levels. Thus they develop a steep lapse rate 
and conditional and convective instability. The condensation level 
is low, and cumulus clouds and showers may develop. 

The degree of instability developed depends upon the length of 
time the air has overlain the ocean. Some of the air masses move 
in a short path across the narrow north Pacific and retain many of 
their cP characteristics. In general they reach the west coast of 
North America as mPk air masses, but as they move inland they be- 
come mPw, because the land surface is cold. The air is cooled at 
the surface and becomes more stable, with little cloudiness and lit- 
tle turbulence. When such air moves against the western mountain 
ranges, however, the orographic uplift results in heavy rain or snow 
on the western slopes of the mountains. 

Some Atlantic air masses originating in the northwestern Atlan- 
tic invade the eastern coastal region of North America from Virginia 
northward, giving raw, northeast winds. They are not numerous 
because of the prevailing west-to-east movement of the air in the 
general circulation. The air on the Atlantic coast differs from the 


air reaching the Pacific coast by being colder near the ground and 
stable aloft. In their lower levels, both have conditional instability 
and high humidity. 

Polar maritime (mP) air masses in summer. The source regions 
of mP air are the same in summer as in winter. Along the west 
coast in summer, there is an almost continuous southward flow of 
air of moderate temperature. There is subsidence and marked sta- 
bility in the dry air aloft, but there is conditional instability in a 
shallow lower layer. This situation results in the low stratus clouds 
and summer fogs characteristic of much of the Pacific coast. This air 
moves inland as mPk air. It is quickly heated at the surface over 
the hot and dry interior, but the upper layers remain dry and stable. 
The heating and the turbulent mixing reduce the relative humidity 
of the lower layer and dissolve the low clouds. The air is dry and 
generally clear, and remains cooler than the surface over which it 
is moving. After it has crossed the Rocky Mountains, this mP air is 
indistinguishable from cP air. 

Along the east coast, air occasionally moves inland out of a high- 
pressure area over the cold water of the northwestern Atlantic. This 
westward movement occurs more frequently in summer than in 
winter because of the change in the general pressure distribution 
with the seasons. The upper air is stable because of subsidence over 
the high-pressure area, and the lower air is cool, dry, and stable be- 
cause of the cold water. It is stable mP or mPw air over the water 
and becomes mPk over land. Because of the stability of the mass as 
a. whole and its low humidity, there are no clouds, or only thin 
stratocumulus, which are dissolved by insolational heating as the 
air proceeds inland. Hence, summer rnP air from the Atlantic often 
brings clear and cool weather and good visibility to the New Eng- 
land states and occasionally to the coastal states as far south as 

Tropical continental (cT) air masses. Because the North Amer- 
ican continent narrows rapidly as it extends southward through 
Mexico into tropical regions, little true cT air ever invades the 
United States. During the winter months there are no cT air masses 
in North America. The only source of such air in the Northern 
Hemisphere in winter is over north Africa. In summer, northern in- 
terior Mexico and adjacent portions of our arid Southwest are in 
the subtropical belt of light winds and light rainfall, and in conse- 


quence, hot and dry masses of air accumulate in those regions. These 
may properly be called cT air masses. 

Owing to the intense heating of the surface, there is turbulence 
and convection to a considerable height (2 miles or 3 km), result- 
ing in a lapse rate approximating the dry adiabatic rate. Notwith- 
standing this steep lapse rate, the air remains cloudless because of 
its extreme dryness. The dryness results also in rapid insolational 
heating by day and rapid radiational cooling by night. Accordingly, 
the weather is hot, dry, and clear, with large diurnal ranges of 
temperature. These air masses are confined to the region of their 
origin. When they move away from this region, they become mixed 
with mT air masses and lose their identity. 

Tropical maritime (mT) air masses in winter. The tropical mari- 
time air that affects the weather of North America has its sources 
in the subtropical high-pressure belt, either in the Pacific Ocean 
between Baja California and Hawaii, or in the Gulf and Caribbean 
regions and the region of the Sargasso Sea. In winter, the tempera- 
ture and humidity of the Pacific mT air masses are moderate. Sub- 
sidence is characteristic of the warm, dry air. The surface turbu- 
lence layer is cool and moist. When this air moves northward into 
colder latitudes, it becomes mTw air. The cooling and convergence 
as it moves northward cause the lower layers to become more stable 
and often cause decreasing visibility, fog, and drizzle, particularly 
at night. These conditions favor convective instability, and when 
such air is forced upward by colder air masses along a front or by 
moving inland and upslope, there is often steady moderate to heavy 
precipitation. These are the conditions under which the winter rains 
of the Pacific coast occur. These mT air masses are greatly modified 
before they cross the continental divide. 

The waters of the Gulf of Mexico and the subtropical western 
Atlantic are exceptionally warm in winter and evaporation from 
their surfaces is active. Hence, they are both warm and humid in 
their lower layers, and thus their specific humidity is unusually high. 
The upper levels of these air masses are like those of tropical Pacific 
air, warm, dry, and stable, owing to subsidence. As they move north- 
ward, there is rapid surface cooling and, hence, condensation in the 
humid lower levels. These happenings produce poor visibility, fog, 
drizzle, and low stratus clouds, resulting in very poor flying weather. 
The air continues warm (mTw), compared to the surface, especially 


over northern land areas, where it often causes winter thaws. The 
upper levels become colder and less stable northward, resulting in 
convective instability and considerable precipitation. When there is 
frontal or orographic uplift, the rains become heavy and wide- 
spread. The mT air masses of the Atlantic and Pacific Oceans have 
essentially the same tropical maritime properties, and, taken to- 
gether, they are responsible for the warm and rainy winter weather 
of the entire central and eastern portions of the United States and 
of southeastern Canada, 

Tropical maritime (mT) air masses in summer. There are prac- 
tically no mT air masses moving over the United States from the 
Pacific Ocean in summer. On the other hand, Atlantic and Gulf air 
masses from the same sources as in winter are more numerous in 
summer and cover a wider area. Temperature, specific humidity, 
and relative humidity are all high in these air masses at their sources 
in the Gulf and western Atlantic. There is conditional instability; 
a small uplift causes strong convective currents and frequent thun- 
dershowers, As they move over land, they become mTk air masses. 
Stratus and stratocumulus clouds frequently form by radiational 
cooling of the humid air at night. These disappear in the forenoon, 
to be followed in the afternoon by cumulus clouds and thunder- 
showers, due to insolational heating of the surface. 

Where there is convergence and uplift, as in frontal zones, the 
mT air becomes definitely unstable aloft, resulting in active con- 
vection and heavy showers or widespread heavy rain. There is little 
change in this air as it moves northward over land, and when it 
crosses the Great Lakes, it is warmer than the water. Fogs and low 
stratus clouds frequently result on the northern shores of the lakes. 
Over the cold water off the northeastern United States and the 
Maritime Provinces of Canada there are deep and dense fogs. These 
air masses largely dominate the summer weather of the eastern half 
of the continent, where they are responsible for much hot, humid, 
and oppressive weather, as well as for most of the rainfall. Most of 
the thunderstorms in the United States occur in mT air. 

Superior (S) air masses. Masses of warm and dry air, known as 
Superior air, are common at middle and upper levels, 6,500 feet 
(2 km) and upward, over most of the United States at all seasons. 
This air is especially common over mT air, but it occurs also over 
air of polar origin. Its extreme drynesss is evidently due to sub- 


sidence. The name Superior air is now often applied to all warm air 
masses having a relative humidity of less than 40 per cent, under 
the assumption that they have become warm and dry by subsidence 
from aloft. The source regions of the warm, dry air aloft are not 
definitely known. It is probable that the greater part of such air 
develops slowly in the upper levels of the subtropical high-pressure 
cells. Much of it reaches North America from the eastern side of 
the Pacific high in great tongues of dry air moving southward out 
of the prevailing westerlies. 

Superior air has a steep lapse rate, approaching the dry adiabatic, 
but it remains stable because of its extreme dryness, It is usually 
warmer at its base than the air which it overlies; that is, there is 
a temperature inversion at its lower boundary, stopping convective 
movements from below. It is essentially a high-level air mass, but 
at times, especially in summer, it appears at the surface, attended 
by hot and dry weather. The Great Plains and the southwestern 
states are subject to such periods of heat and drought, due to the 
continued presence of S air at or near the surface. It is present much 
of the time at all seasons over wT air in central and southern United 

Formation and Characteristics of Fronts 

A front is a boundary surface, or, more correctly, a transition zone, 
separating air masses of differing character, specifically of markedly 
different temperatures. It is a sloping boundary and comparatively 
narrow, varying from 50 to 500 miles in width. 

When differing air masses are brought together by converging 
movements in the general circulation, they ordinarily do not mix 
freely but form a transition zone, across which there is a rapid 
change in temperature. The cold air underlies the warm air in a 
sloping, wedgelike mass. The natural tendency is for the warm air 
to lie above the colder, denser air in a horizontal layer, but the 
continuous forces of pressure differences and the earth's rotation 
never permit this state of equilibrium to be reached. The front 
shown on a weather map is the line along which an inclined bound- 
ary surface between two air masses reaches the ground. A wedge- 
shaped cold air mass is shown invading a region of warmer air in 
the three-dimensional sketch of Fig. 82. 




Fig. 82. Three-Dimensional Representation of a Cold Front. 

Frontogenesis, frontolysis, cyclogenesis. The formation of new 
fronts, or the regeneration and strengthening of weak and decaying 
fronts, is called ft 'ontogenesis. The opposite process, that of the 
weakening or dissipation of existing fronts, is frontolysis. Fronto- 
genesis occurs where the wind system causes a convergence of cold 
polar air and warm tropical air. A strong contrast in temperature is 
a necessary condition, and any process that causes and maintains 
an increased temperature gradient, a crowding of the isotherms 
closer together, tends to produce a front. Similarly, any distribution 
of winds that causes divergence of air masses and separation of iso- 
therms is a process of frontolysis. 

The two most active regions of frontogenesis in the Northern 
Hemisphere in winter are: (1) the north Atlantic Ocean from thr 
region of the Iceland low westward to the coast of North America, 
and (2) the north Pacific from the region of the Aleutian low west- 
ward to the coast of Asia. In summer there is active frontal devel- 
opment in the Bering Sea region and across central Canada. These 
are regions in which the zonal and cellular wind systems bring 
together air masses of strongly contrasting characteristics. 

The regions just mentioned are the principal breeding places of 
the migratory extratropical cyclones, to be discussed in the next 
chapter, which are so large a factor in the weather of the zone of 
the prevailing westerlies. The formation of such a storm on a front 
is called cyclogenesis. Changes in the wind direction and force or 
in the temperature distribution along a front, if of sufficient magni- 
tude and if the warm air is on the forward side, result in the forma- 
tion of a frontal wave having a cyclonic circulation. Such a wave 



develops into an active cyclonic storm, provided ( 1 ) that the dis- 
turbing influence is strong enough to produce a wave of consider- 
able size, and (2) that there is a marked temperature contrast be- 
tween the two air masses. Irregularities of flow, such as are required 
to initiate waves, are to be expected in all wind currents. Larger 
wind disturbances often result from the approach of secondary 
fronts and of pressure waves aloft. The regions of most active f ronto- 
genesis are also, as would be expected, the regions of most active 

Characteristics of warm fronts. At a warm front, warm air is ad- 
vancing against cold air and being forced upward over an under- 
lying wedge of the cold air. The resulting adiabatic cooling of the 
warm air takes place slowly because the upslope is gentle. The 



Fig. 83. Surface and Vertical Section of a Warm Front. 


amount and type of cloudiness and precipitation resulting from this 
upward movement depend upon the existing humidity and lapse 
rate in the warm air. If the air is stable and dry, there may be lit- 
tle cloudiness and no precipitation. In most cases in the United 
States, however, the warm air is a tropical maritime air mass, either 
from the Pacific or from the Gulf and Atlantic source regions. Such 
air is humid and normally conditionally and convectively unstable. 
Hence, the initial uplift usually leads also to convective ascent. In 
advance of an approaching warm front of this character, we find 
first a slowly falling barometer and the formation of high clouds. 

The clouds may begin as much as 1,000 miles in advance of the 
surface front. They are cirrus clouds, thickening into cirrostratus 
(Fig. 83). As the front approaches, cloudiness begins at interme- 
diate levels, the sky becoming overcast with altostratus and alto- 
cumulus forms. Slow, steady rain may begin from the altostratus. 
The temperature is constant or slowly rising, unless lowered by 
falling rain. As the front draws near, there is an increasing fall in 
pressure, and the clouds lower and thicken into stratocumulus and 
nimbostratus, attended by steady, moderate rain or snow. Some- 
times there is sufficient instability to produce cumulonimbus clouds 
and thunderstorms. As the warm rain falls through the underlying 
cold air, evaporation of the raindrops, combined with turbulent 
movements of the lower air, may result in low stratus clouds and 
fog. This condition creates a serious hazard to aviators. In winter, 
there may also be icing from supercooled drops. With the passage 
of the front, there is normally a gradual rise in temperature, a 
change in wind direction, and generally clearing weather, although 
some cloudiness may continue throughout the warm-sector air mass. 

Characteristics of cold fronts. In a cold front, warm air is being 
replaced by an advancing wedge of cold air. As in the case of the 
warm front, the vertical structure of the warm air determines the 
reactions with reference to cloudiness and rainfall. A cold front 
differs from a warm front in the following particulars: (1) It is 
steeper, giving the same uplift in a shorter distance. (2) It slopes 
backward instead of forward. (3) The warm air is being removed. 
The following characteristics of cold fronts result: ( 1 ) There is usu- 
ally no warning far in advance of an approaching cold front unless 
thunderstorms are especially active along the front. In the latter 
case, cirrus or thin cirrostratus from the anvil tops of the cumu- 



lonimbus may be carried by strong westerlies aloft and precede the 
front by several hours. (2) There is only a narrow band of cloudi- 
ness and precipitation. (3) The reactions are sharper and more 

As the typical active cold front draws near, there is some increase 
of wind in the warm sector and cirrus or cirrostratus clouds appear. 
These are quickly followed by lower and denser altocumulus and 
altostratus cloud forms, and then at the actual front by nimbo- 
stratus and cumulonimbus, with heavy showers. These changes take 
place within an hour or two. As the front passes, there is a rise in 
pressure, an abrupt and often large drop in temperature, an increase 
in wind force, and a change in direction from southwesterly to 
northwesterly. These events are usually followed by a fairly rapid 
clearing, but scattered cumulus or stratocumulus may persist for 
some time (Fig. 84). 

There are, of course, many individual variations from the typical 



Fig. 84. Surface and Vertical Section of a Cold Front. 


conditions, depending upon the characteristics of both air masses, 
hence the importance of upper-air soundings in analyzing an indi- 
vidual front. Sometimes most of the rain comes just ahead of the 
front; sometimes, a little behind it. If the front moves very slowly, 
or if the slope of the frontal surface is not very steep, clouds and 
precipitation extend backward a considerable distance. 

When the cold air moves over a warm surface, especially a warm 
water surface, evaporation often produces low clouds or fog. In 
other cases, the falling raindrops freeze, forming sleet, or they be- 
come supercooled and deposit a layer of ice when they strike sur- 
face objects, When the cold front is moving rapidly and increasing 
its speed, one or two secondary cold fronts sometimes develop some 
distance behind the main front. 

Along the cold front of an advancing high, where polar air quickly 
replaces tropical air, there is a sudden change from warm to much 
colder weather. This occurs especially in winter in interior and 
eastern North America and is known as a cold wave. As a great drop 
in temperature advances rapidly eastward and southward, its move- 
ment suggests an oncoming flood or large ocean wave. The defini- 
tion of a cold wave varies with the season and the locality. Cold- 
wave warnings are issued as far in advance as practicable in order 
that people may prepare for these sudden changes. 

Squall lines. One of the most respected features on the weather 
map and one of the most difficult to predict is the pre-cold-frontal 
squall line. It is characterized by a line of heavy showers and thun- 
derstorms aligned parallel with the cold front and advancing ahead 
of it. 

Squall lines are commonly identified with cold fronts moving 
across the United States in the spring but may occur during ail 
seasons. Storminess along a squall line intensifies and dissipates more 
or less erratically. The line forms from 50 to 150 miles (80 to 240 
km ) in advance of a cold front which is pushing into warm, condi- 
tionally unstable air. Once formed, a squall line usually travels faster 
than the parent front and dissipates when the distance between 
them approaches 200 or 300 miles (320 to 480 km). When con- 
vective activity along the squall line increases, there is a correspond- 
ing decrease in activity along the main front, thus indicating that 
the dynamics of the two lines of weather are very closely related. 


The exact cause of pre-cold-frontal squall lines is not known. One 
theory holds that outnishing cold air from the downdrafts of thun- 
derstorms along the main front creates a secondary frontal situa- 
tion in the warm air mass. Another advocates that surface friction 
may retard the advance of cold air near the ground, causing an 
almost vertical frontal surface at lower levels. The free movement 
of warm air up the frontal siirface would thus be impaired and re- 
sult in a band of convection ahead of the frontal surface, Convection 
is present along a squall line, as most any pilot can testify, some- 
times creating weather conditions more commonly associated with 
a very active cold front. 

Other types of fronts. Basically, all fronts are either warm or 
cold; but when a front ceases to move in either direction, it is called 
a stationary front. This is not an uncommon occurrence, especially 
along the eastern edge of the Rocky Mountains where the physical 
barrier hinders free movement of the air masses. Cold fronts often 
become stationary along the Gulf Coast and across Texas or Mexico 
as the cold air ceases to advance further southward and the main 
body of cold air moves off to the east. A front may remain stationary 
for several hours or days before it dissipates or begins to move again 
as a cold front or a warm front. 

It sometimes happens that one front overtakes another front, 
bringing into close proximity three different air masses. This situa- 
tion becomes possible when the middle air mass is warmer (and 
lighter) than the other two. As the fronts meet, the air masses are 
displaced vertically according to their densities. The wannest air 
mass is completely occluded from the ground and the frontal struc- 
ture becomes known as an occluded front or an occlusion. Occluded 
fronts are generally associated with wave cyclones and will be dis- 
cussed more fully in Chapter 10. In an occlusion, one of the fronts 
no longer touches the ground. Its lower boundary terminates along 
the frontal surface which separates the coldest air from the other 
two air masses. A front that does not reach the ground is called an 
upper front. 


1. Assuming a cold front slope of Vrs* how high would be the frontal 
surface over Oklahoma City when the surface front extended through 
St. Louis, Little Rock, and Dallas? 


2. From a series of United States Weather Bureau maps: 

(a) Determine the movement of air masses from day to day. 

(b) Classify the air masses according to their source regions and 
the areas over which they pass. 

(c) Determine how much the air masses are modified from day to 
day as indicated by changes in temperature and humidity at 
the surface. 

(d) Follow the daily movement and modification of the various 

(e) Note the state of the weather within the various air masses and 
on both sides of the various fronts. 

(f) If series of weather maps for both summer and winter are 
available, study the variation and intensity of air-mass and 
frontal activity during the winter months as compared to the 
summer months. 




The general circulation and the movement of air masses, as dis- 
cussed in the preceding chapters, may conveniently he regarded as 
a background upon which are superimposed many smaller disturb- 
ances and irregularities. It is like the flow of a river, with many 
eddies and cross currents. Some of these lesser movements, such as 
sea and land breezes and katabatic winds, have been noted. The 
irregularities to be considered here are of quite another type, to 
which the name secondary circulation is particularly applicable. 
These are traveling disturbances, of which some originate in high 
latitudes and others in the tropics. They are closely associated with 
air-mass movements and frontal activity in causing the day-by-day 
weather changes in the temperate latitudes. 

Little was known about the characteristics and behavior of mov- 
ing wind systems until about 100 years ago. Before discussing them 
here, it seems advisable to note the device by which much of our 
present knowledge about them has been gained. 

Weather Maps 

As the techniques of rapid communication were improved, it be- 
came more practical to try to produce a composite graphical pic- 
ture of weather conditions at periodic intervals. Scientists in Ger- 
many, England, and America spent much time before 1850 trying 
to accumulate simultaneous observations of weather conditions over 
a wide area. Before these could be collected and entered on a map 
of the area, several days had passed and weather changes of major 



proportions had occurred. Soon after the use of the telegraph 
became widespread, simultaneous weather observations could be 
quickly collected and charted, permitting inferences to be drawn 
about future changes in the weather. This was the beginning of 
weather forecasting on a scientific basis. 

Synoptic charts. A synoptic weather chart is designed to present 
a view of the whole weather situation at one time. Weather obser- 
vations are made at selected stations over the area to be represented. 
This information is collected by teletype and plotted on the chart 
at each station location. Analysis lines and symbols are drawn show- 
ing the distribution of pressure, temperature, air masses, fronts, pre- 
cipitation areas, and the like. ( See Figs. 86, 87, and 88. ) On a com- 
plete surface chart, data are also entered in station-model form 
showing amounts and types of clouds, wind direction and velocity, 
visibility, pressure changes, and so forth (Fig. 85). Such maps are 
usually prepared four times daily. Once a student has followed the 
progression of weather features through several map intervals, he 
will surely become aware that the synoptic weather chart is a pow- 
erful forecasting tool. Charts may also be drawn for levels other 
than the surface when upper-air data are available. Such charts are 
usually prepared for the 850-, 700-, 500-, and 300-mb surfaces. They 
do not show all the details of the surface chart, but they give a 
good picture of the distribution of pressure, temperature, humidity, 
and wind. 

Fig. 85. Station Model for Plotting Synoptic Weather Data. The information is 
interpreted as follows: wind, northwest, Beaufort force 4; temperature, 67 F.; distant 
lightning; visibility, 10 miles; clew point, 65F.; towering cumulus with bases at 1,000 
feet and altostratus; pressure, 1001.7 millibars, having risen 0.4 millibars in an un- 
steady fashion during the last three hours; there has been a thunderstorm at the 
station during the past six hours. 

Making a weather map. The making of a weather map requires 
the work of a large organization. First, there must be, over a large 
area, a well-distributed network of meteorological stations at which 


competent, trained observers make synchronous observations of the 
weather. Second, the data so obtained must be collected rapidly by 
wire or radio at the centers where maps are to be prepared. Third, 
the information so collected must be quickly mapped to show the 
distribution and intensity of the weather elements. Fourth, from this 
picture of the weather, the forecaster must make his interpretations 
and inferences. Finally, the forecasts must be promptly distributed 
if they are to serve their purpose of informing the public and giving 
the news of the weather to those interested. 

Value of weather maps. The entire system of short-period 
weather forecasting is dependent on the synoptic chart and the sup- 
plementary charts. From the picture of the existing weather thus 
set before him, the meteorologist is able to estimate with fair ac- 
curacy the changes that will occur in a given area during the next 
24 to 48 hours. He can do this for a distant area about as well as 
for his own locality if he has become thoroughly familiar with the 
behavior of the weather in the area. This is true because he makes 
his forecast primarily from a study of the map. His inferences are 
based partly upon known physical laws governing the behavior of 
the atmosphere and partly upon familiarity with previous maps, that 
is, upon a knowledge of how the weather has behaved before under 
similar conditions. 

During the time that weather maps have been available, they 
have been carefully examined, studied, and classified by many stu- 
dents. They have been proven indispensable in the making of 
weather forecasts and in the scientific study of many meteorological 
problems. They have not, however, completely fulfilled early ex- 
pectations. They have not led to the explanation of all the phe- 
nomena of the air nor to the development of perfect weather fore- 
casting. Much has been accomplished by their aid, but much re- 
mains to be done before a complete understanding of the atmos- 
phere is reached. 

Low-Pressure Centers 

If we examine even a short series of weather maps, we find that 
the isobars do not have the same regularity of spacing and of direc- 
tion that is shown on the charts of the general circulation. Instead, 
the isobars are disturbed by irregularities of pressure and assume 




various shapes and patterns which change their locations and alter 
their shapes and patterns from day to day. Two patterns most easily 
recognized and of great importance in understanding the weather 
are those which enclose areas of low or high barometric pressure. A 
low-pressure center which is enclosed by one or more isobars is 
often called a cyclone or cyclonic depression. 

Characteristics of cyclones. The cyclone is a barometric depres- 
sion marked by a series of roughly circular or oval isobars inclosing 
an area of low pressure, that is, where pressure decreases from its 
outer rim to its center. ( See Fig. 89. ) Such a system has long been 
known in meteorology as a cyclone or an extratropical cyclone. The 
word cyclone carries the idea of a revolving storm. The names de- 
pression, cyclonic depression, and low seem preferable, because it 
is now known that such a traveling disturbance is not always com- 
posed of a revolving mass of air, and also because the name cyclone 
has been applied to storms of a different nature (the tropical dis- 
turbance and the tornado, to be noted later). However, the name 
cyclone is well established to designate the type of pressure pattern 
discussed in this section. 

Individual cyclones differ greatly 
in size, ranging in diameter from 
100 to 2,000 miles, the average di- 
ameter in the United States being 
1,000 miles or more. They also vary 
in form from approximate circles 
to much elongated ovals. The ovals 
are sometimes so much flattened at 
one end as to receive the name of 
V-shaped depressions and some- 
times become so broad and shallow 
that they are called troughs of low 
pressure. When thus greatly elon- 
gated, they lose some of the fea- 
tures generally regarded as char- 
acteristic of cyclones. The typical round or elliptical low has, near 
the surface of the earth, moderate winds directed inward and around 
the center of low pressure, making angles of from 20 to 40 with 
the isobars. The direction of movement is counterclockwise in the 
Northern Hemisphere, responding to the influence of the earth's 


Fig, 89. Barometric Depression 
Showing Cyclonic Circulation in the 
Northern Hemisphere. 


rotation and following Buys-Ballot's law. Such a movement of air 
around a center of low pressure is called a cyclonic circulation. ( See 
Fig. 89.) Cloudiness and precipitation are usually associated with 
a cyclone. 

Source regions. Cyclones, or low-pressure systems, are associ- 
ated with the equatorward, shifting movements of the polar front 
and usually originate outside of the tropics, more frequently in high 
latitudes. They are more numerous and better developed in winter 
than in summer. In the Northern Hemisphere, many begin in the 
North Pacific and the North Atlantic Oceans most frequently in 
the regions extending from their western borders eastward to the 
Aleutian and Iceland lows. Many have their origin in the region of 
China and the Philippines. Others that affect the United States 
begin in western Canada, or in western and southern portions of 
the United States. In the United States, lows are given names in- 
dicating their place of origin or first appearance on the weather 
map: Alberta, north Pacific, south Pacific, northern Rocky Moun- 
tain, Colorado, Texas, east Gulf, south Atlantic, and central. Of 
these, the Alberta low is the most frequent, and those that come 
from the east Gulf and South Atlantic regions are the least numer- 
ous. Lows from the different regions have somewhat different char- 
acteristics and paths. A secondary low, having the same general 
characteristics, frequently develops on the equatorward side of the 
primary depression. This is especially likely to occur in elongated- 
oval or V-shaped depressions. 

Movement of cyclones. The general direction of motion is from 
west to east, with frequent trends to southeast or northeast. Each 
individual depression, in fact, seems to select its own path as it 
makes its way eastward. There is no fixed path which all follow, but 
there are general tracks more frequented than others. The lows 
originating in the western Pacific move northeastward by way of 
Japan and the Kurile Islands to the Bay of Alaska. From there they 
move southeastward, as do those that have their beginning in the 
Aleutian low, to enter the continent as north Pacific or Alberta lows. 
Across North America, there are three predominant paths: (1) east- 
ward along the border of the United States and Canada; (2) from 
western Canada or the north Pacific southeastward into the Missis- 
sippi Valley, and thence northeastward to the Great Lakes, the New 
England states, or the St. Lawrence Valley; (3) from the south- 



western region eastward to the Mississippi Valley and then north- 
eastward to New England. Typical paths of the nine named types 
are shown in Fig. 90. 

Fig. 90. Typical Paths of Cyclones Appearing in Various Regions of the 
United States. After Bowie and Wcightman. 

Some of the cyclones from North America cross the Atlantic 
Ocean to Europe. Most of these, together with those that begin in 
the north Atlantic, move northeastward across or north of the British 
Isles into Russia. Some curve farther south and enter Europe by way 
of France. The rates at which the cyclones advance are variable 
and individual, like their paths. The average movement is 20 to 30 
miles (30 to 40 km) per hour. The higher averages occur in winter 
and the lower in summer. 

By thus giving different names to the cyclones originating in dif- 
ferent regions, and by following individual disturbances for con- 
siderable distances, we seem to treat them as independent entities; 
but the fact should be kept in mind that they are comparatively 
small irregularities in the larger, more orderly movements of the 
air. They are not independent of the general circulation but are in- 
terruptions in its symmetry. They are found along the boundaries 
(fronts) between adjacent air masses and move somewhat as the 
air masses move. The air in a low-pressure center usually consists of 
the air along the edges of two or three air masses in close proximity 



to each other. In this case, the different air masses are separated 
from each other by frontal surfaces. The apparent circular wind 
flow, characteristic of moving cyclones, is the result of separate 
wind movements in each of the contributing air masses. 

High-Pressure Centers 

The other characteristic pattern of isobars to be observed on al- 
most any weather map is the high-pressure area or anticyclone, 
sometimes simply called a "high." 

Characteristics of anticyclones. Isobars enclose an anticyclone 
in circular or elliptical fashion. Winds spiral outward around the 
center, clockwise in the Northern Hemisphere, gradually crossing 
the isobars toward low pressure. This system of diverging winds 
constitutes an anticyclonic circulation, illustrated in Fig. 91. The 
area within the closed isobars is often larger than in a cyclone, and 
the pressure gradient is smaller (the isobars are farther apart). 
Winds are generally light in conformity with the pressure gradient, 
and calms are common near the center. There is usually little cloudi- 
ness, although heavy clouds and precipitation may characterize the 
advancing edge of a moving anticyclone in the area near the frontal 
boundary. In the Northern Hemisphere, the eastern half of a travel- 
ing high is relatively cold at the surface with northerly winds, and 
the western half is warm with southerly winds. 

Unlike a low-pressure cen- 
ter, which may be composed 
of two or more air masses, an 
anticyclone usually consists of 
a single air mass with more or 
less homogeneous properties. 
Identifying characteristics of 
temperature and humidity can 
be traced to the life cycle of 
the circulation, including its 
source region and trajectory. 
Tracks and velocities. The 
movement of anticyclones is 
similar in general to that of 
cyclones, highs and lows often 


Fig. 91. Anticyclone and Clockwise Circu- 
lation in the Northern Hemisphere. 

following each other in regular succession. This is particularly true 



iii middle latitudes of the Southern Hemisphere, where the surface 
is largely a water surface. There, the regularity is such as to form a 
wavelike procession around the earth. Over the large land areas of 
the Northern Hemisphere, highs are more likely to become station- 
ary, or nearly so, than are lows, and their progress sometimes comes 
to resemble spreading rather than traveling. They then become iso- 
lated areas of high pressure with lows moving past them on either 
side. Such stagnation occurs more frequently in Europe than in 

Fig. 92. Typical Paths of Anticyclones Appearing in Various Regions of 
the United States. (After Bowie and Weightman.) 

The principal types of American highs, named according to their 
source regions are: Alberta, north Pacific, south Pacific, Plateau and 
Rocky Mountain, and Hudson Bay. It will be noted in Fig. 92 that 
their average paths differ somewhat from those of cyclones, and 
in particular that they enter the Atlantic Ocean farther south. The 
cyclones appear to be attracted by the Iceland low; the anti-cyclones 
remain south of it. 

Nature and Origin of Highs and Lows 

In connection with the formation and maintenance of cyclones 
and anticyclones, two primary facts as to their nature are to be kept 


in mind: (1) In a low, air is moving inward toward a center, and 
some of it is being carried up and removed at the top; (2) in a high, 
air is added at the top and is slowly flowing out at the bottom. Any 
theory of their origin must account for these facts and for the sup- 
ply of energy which maintains and moves them. 

Convection theory. Early attempts to explain the genesis of cy- 
clones were along the lines of convection. It was assumed that a 
low was a mass of warm, moist air, rotating around a center where 
the air was rising by thermal convection, and that the overflow 
formed the adjacent high. Further knowledge of the actual condi- 
tions obtaining in lows and highs shows that this explanation is in- 
adequate and not in agreement with the facts. Some of the reasons 
for the rejection of this theory are: ( 1 ) Lows are more frequent and 
better developed in winter, when convection is less active, than in 
summer. (2) They frequently begin over oceans, where surface 
heating is negligible. (3) The depression or low-pressure system 
as a whole usually is not a single body of warm air revolving around 
a center. 

These facts seem to exclude local heating of the surface air as a 
primary cause of cyclonic depressions. Brunt has pointed out, how- 
ever, certain special conditions under which the origin of a moving 
low may be ascribed to surface warming. When a mass of cold air 
moves from polar regions into warmer latitudes and passes over 
progressively warmer land or ocean surfaces, the lower layers are 
wanned. Thus convection may begin over a large area. If the rising 
air is soon saturated and if the lapse rate is between the dry and 
the saturated adiabatic rates, the upward motion thus begun may 
continue to great heights and over a large enough area to initiate a 
characteristic cyclone. It should be noted, also, that stationary areas 
of low pressure frequently form over heated regions, as in Arizom 
in summer, and are properly called heat or convection lows but may 
not have cyclonic circulations. 

Polar-front theory. Originating with the Norwegian meteorolo- 
gist, V. Bjerknes, 1 a more definite explanation of the origin of cy- 
clones and anticyclones has been developed since about 1915. In- 

1 V. Bjerkncs, "On the dynamics of the circular vortex with applications to the at- 
mosphere and atmospheric vortex- and wave-motions." Kristiania: Geofysiske Publi- 
kationer, Vol. II, No, 4, 1921. 


stead of a gradual, uniform change of temperature from equatorial 
to polar regions, Bjerknes envisages masses of cold air accumulated 
in polar regions and masses of warm air in equatorial and tropical 
regions. In the region of the prevailing westerlies, these masses of 
cold and warm air meet and thereby form a surface of discontinuity , 
a well-marked and distinct surface of separation between the two 
masses. This surface is the polar front (see Chapter 9), across which 
there is a sudden change in the temperature of the air and often in 
its humidity. Irregularities of flow along the polar front are thought 
to initiate depressions, and the energy of flow of the two masses of 
air combines with the instability due to their difference in density to 
develop and maintain them. 

The essential condition for the formation of a cyclone is the exist- 
ence of bodies of warm and cold air adjacent to each other. The 
assumption of some such meeting of large masses of air of different 
temperatures and humidities and moving in different directions is 
now generally accepted as affording the best available basis for the 
study and interpretation of the weather phenomena attending 

Idealized wave cyclone. Fig. 93 indicates the process of develop- 
ment of a cyclonic depression according to the wave theory. We as- 
sume, first, a current of warm air moving from the west on the south 
side of a cold current moving from the east, the two currents being 
separated by a definite surface of discontinuity, or front. Then some 
local disturbance or irregularity of movement causes the turning of 
the warm air toward the cold current. This sets up a wave in the 
front, as shown in Fig. 93 (a), and divides it into two parts, The east 
portion of the discontinuity, shown by the double line, where warm 
air is advancing eastward and meeting colder air, is now the warm 
front. The west portion ( single line ) , where cold air is replacing the 
warm air, is the cold front. Sometimes the wave thus started moves 
along the polar front without further development, attended by rain 
in the region indicated by the stippled area in (a), and ultimately 
dying out. More often, the curving motion having been begun, the 
movement of the two currents and the deflective force of the earth's 
rotation result in additional curvature and increased amplitude of 
the wave, as in ( b ) . This represents a fully developed young depres- 
sion composed of the original streams of warm and cold air, but now 




' " > Hfr/-/7? Air i Warm Front 

*- C*/</ Air Co/rf /ro/7/- 

Occluded Front 

Fig. 93. Stages in the Development of a Traveling Wave Cyclone. 

divided into warm and cold sectors. The cold front was recognized 
many years before the development of the polar-front theory and 
was named the wind-shift line or squall line. 

Along the warm front, the warm air overtakes and overrides the 
colder air; and along the cold front, cold air is overtaking warm air 
and forcing it up. (See the vertical cross section of Fig. 93.) There 
is, accordingly, rising air along the entire frontal surface; conse- 
quently, if the air is moderately moist, there is cloudiness and pre- 
cipitation along both warm and cold fronts, as shown by the stipled 
areas in (b) and (c). 

Occlusion of a wave cyclone. By the process just described, the 
warm air is being displaced at the surface of the earth by each of the 
flanking air masses. When the cold front overtakes the warm front 
and cuts off the warm air at the surface, the cyclone is said to be 
occluded. ( See Fig. 93 ( d ) . ) The two fronts merge into one occluded 
front, where cold air from the west meets cold air from the east. 
Although these two polar air masses may have been alike in their 


origin, each has undergone modifications according to its history 
and path since leaving the source region. When they meet, they 
probably differ considerably in temperature and moisture, and the 
lighter, wanner one is forced up by the colder. Two distinct types of 
occlusion are recognized. If the air advancing as the original cold 
front is colder than the air forming the wedge ahead of the original 
warm front, the occlusion is of the cold-front type. The cold front 
continues at the ground level, but the warm front is displaced up- 
ward along the cold frontal surface. The front that has been removed 
from the ground is sometimes referred to as an upper front. In this 
case, the more common situation, it is called an upper warm front. 
( See Fig. 94. ) 

In some cases, however, the wedge of cold air ahead of the warm 
front is colder than the advancing wedge of cold air behind the cold 
front, thus forming a warm-front type of occlusion. The warm front 
continues in contact with the ground and the two warmer air masses 
are forced up the warm frontal surface (Fig. 94). Now the upper 
front is an upper cold front. In both cases, two air masses are forced 
up a common frontal surface, and the warmest air is steered up a 
trough created by the contact of the* two frontal surfaces. Cloudiness 
and rainfall are typical in the area of an occlusion, for, in reality, the 
weather of a cold front and that of a warm front are combined. But 
the characteristics of these fronts arc less definite than those of 
simple warm and cold fronts, because of the great variability of their 
structure. Upper fronts also occur without any connection with a 
surface occlusion, and sometimes with pronounced effects on the 

Secondary low-pressure centers. It is evident that, when an oc- 
clusion has occurred, the cyclone is beginning to die. The warm sec- 
tor is being reduced and the low is filling up and diminishing in 
intensity, since contrasting streams of warm and cold air are essen- 
tial conditions of a cyclone. When the development reaches the 
stage shown in Fig. 93 (d), the original low-pressure system begins 
to die out. Under such conditions, a secondary low frequently de- 
velops south and west of the original one (in the Northern Hemi- 
sphere). There may be a family of cyclones, each occurring some- 
what h rt Mnd the previous one, and each going through a somewhat 
similar process of development, vigorous activity, and decay. 

Vertical cross section of a cyclone. The vertical cross section 








Fig. 94. Cold Front ( A ) and Warm Front ( B ) Types of Occlusion. 

along the line AB, Fig. 93 (/;), of the young and active depression 
shows how the warm air rises over the cold air along the warm front, 
and how the cold air in the rear overtakes and pushes up the warm 
air. The slopes of the surfaces of discontinuity are exaggerated in the 
figure, the actual slope being usually between '.-,0 and Via in the 
case of the cold front and between Moo and !<oo for the warm 
front. Along the warm front, the air rises slowly and gradually, east- 
ward from where it touches the ground, producing a broad belt of 
cloudiness and rain, often in the form of continuous sheets of cloud 
and slow, steady rain. Along the cold front, the surface air is re- 
tarded by turbulence and the front becomes nearly vertical for a 
short distance above the ground. This produces an abrupt and vig- 
orous upward motion of the warm air, resulting in the rapid devel- 
opment of cumulus and cumulonimbus clouds, attended by squally 
weather, often with showers and thunderstorms. This band of cloud- 
iness and showers is usually narrow, and is soon followed by the 
clear, cooler, drier air typical of the cold-air mass. 

Movement of air in a cyclone. Highs and lows move eastward, 
carried onward by the prevailing westerlies, often maintaining their 
identities for considerable periods, They can be followed and recog- 
nized from one map to the next even though the shape and arrange- 
ment of the isobars undergo marked variations. A few have been 


followed entirely around the globe. What is it that thus travels? It is 
not the same body of air, for Shaw has shown that the actual move- 
ments of given masses of surface air in a traveling depression are 
quite variable. Some of the air is carried along by the moving de- 
pression and finally reaches its center and rises. Much of it falls 
behind and never arrives at the center but moves off in other direc- 
tions, especially toward the equator. None of it moves entirely 
around the depression. 

Origin of anticyclones. An examination of the paths of typical 
highs originating in Canada and the United States shows that most 
of them follow closely in the rear of well-developed lows. The active 
circulation engendered by the depression, with northerly winds on 
its western side, brings down a surge of cold air from northerly 
regions. Because the pressure is high in this cold, dense air, the sur- 
face air is forced out at the bottom, thus starting an anticyclonic 

At other times we may have a single large mass of cold air moving 
southward, irrespective of the presence of a preceding low. In polar 
regions, in the area between the prevailing westerlies and the polar 
easterlies, there are often large masses of air having little movement, 
This quiet air becomes continually colder by radiating its heat to the 
cold earth and the clear skies. Consequently, it settles, the isobaric 
surfaces bend downward, other air moves in above it, and the pres- 
sure increases. When the pressure becomes too great, a portion of 
this cold air breaks loose and flows southward, forming a tongue or 
wedge of cold air protruding into the warm westerly winds. (See 
Fig. 95.) This puts an obstruction across the flow of the westerlies, 
reducing the pressure on the eastern side of the wedge, forming a 
depression there, and piling up the air on its western side to form 
an anticyclone. 

Two types of anticyclones. Highs crossing the United States and 
Canada begin as surges or wedges moving southward from polar 
regions and may be regarded as moving masses of cold air in the 
lower troposphere. Their depth is ordinarily only 1 or 2 miles (1%- 
3 km) but is sometimes greater. In winter there is frequently a stag- 
nant high over the plateau region of western United States, but the 
typical anticyclone of North America is* a shallow, moving high. 
Such highs are the result of the presence of cold air near the surface 



of the earth. Hence, they are called cold or shallow anticyclones. 
There is subsidence in the cold air with divergence near the surface 
and inflowing warmer air aloft. 

Sometimes, the highs originate as shallow anticyclones and then 
develop into large and relatively warm high-pressure areas. These 
are known as warm or deep anticyclones and more nearly resemble 
the semipermanent high-pressure systems of the subtropics. Deep 
anticyclones are not as mobile as the shallow type, and when one 
becomes established over a continental area in summer, prolonged 
"heat-wave" and "drought" conditions may result. 

Atmospheric Circulation in the Tropics 

Unlike the oscillating surges of 
pressure with the frequent passage 
of highs and lows in middle lati- 
tudes, the pressure pattern of the 
tropics is more persistent. A low- 
pressure belt, or equatorial trough, 
extends latitudinally around the 
earth and is flanked on either side 
by semipermanent high-pressure 
belts, or subtropical ridges, with 
characteristics similarly persistent. 
The equatorial trough and the sub- 
tropical ridges migrate north and 
south with the change of seasons, 
the former shifting through 20 de- 
grees of latitude while the ridges shift only 5 or 6 degrees. 

Pressure, winds, and weather. The mean pressure of the equa- 
torial trough is from about 1010 to 1012 mb, with a very flat pres- 
sure gradient increasing gradually toward the subtropical ridges 
(Fig. 96). Isobars trend east-west, but may meander considerably 
owing to the flatness of the pressure field or to tropical disturbances. 
Diurnal variations in pressure at a given station may exceed the 
variations over a much longer period of time. As a result, meteorolo- 
gists have found isobars to be a less useful tool for weather analysis 
in the tropics than in higher latitudes. 

Streamlines are used in tropical weather analysis to examine the 

Fig. 95, Wedge of Cold Air Invad- 
ing a Warm Current. A cyclonic cir- 
culation is beginning on the east, and 
an anticyclonic circulation on the west. 










Fig. 96. Pressure Profile of the Tropics. Drawing by Vem Askew, 

horizontal wind field. To a degree, they will also represent the pres- 
sure distribution; but, since streamlines have no numerical values, 
they can give no quantitative measure of pressure. They are drawn 
parallel to the wind direction everywhere and give a qualitative 
picture of wind direction and speed (Fig. 97). Like most other 
weather charts, a reliable streamline analysis requires a fairly dense 
network of reporting stations. Lack of data is one of the principal 
handicaps of the synoptic meteorologist in the tropics. 

SEPT. 14, 1953 

Fig. 97. Streamline Chart of the Caribbean Area. After Herbert Riehl. 

The equatorial trough is flanked by the easterly trade winds which 
converge from the northeast and the southeast. (See Fig. 77.) This 
converging air convectively rises above the equatorial trough; other- 


wise, the trough would necessarily fill. The convection gives rise to 
cumulus clouds and thunder showers, which are typical of that 

Convergence of the trades from the two hemispheres often sets up 
one or more lines of squalls or thunderstorms which many meteorol- 
ogists have called an intertropical front. This concept was especially 
prevalent during World War II. It is more commonly agreed today, 
however, that a front in the true sense seldom exists in the equatorial 
trough because the two converging air streams usually lack the nec- 
essary temperature contrast. 2 Furthermore, the bands of weather 
sometimes shift erratically, jump, or completely disappear in a 
fashion that is not characteristic of frontal activity. A more satis- 
factory name for this equatorial phenomenon is intertropical con- 
vergence (ITC). 

Easterly waves. Waves have been recognized and studied in the 
westerlies of middle latitudes for many years. The easterly wave is a 
more recently discovered phenomenon of the trade winds. It may 
be detected by an area of squally weather, a poleward bulge in the 
surface isobars or streamlines (Fig, 97), or by careful analysis of 
the upper-air data. 

The cause of the easterly wave is somewhat obscure, but undoubt- 
edly it is related to perturbations or large eddies in the general cir- 
culation of the atmosphere. Once developed, an easterly wave tends 
to perpetuate itself and move along the equatorward boundary of 
the subtropical ridge from east to west. The typical wave is about 15 
degrees of longitude in length, has an amplitude of from 1 to 5 de- 
grees of latitude, and moves toward the west at from 10 to 20 miles 
per hour ( 15 to 30 km per hour). It brings cloudiness and precipita- 
tion to the area over which it passes and sometimes moves out of 
the tropics by curving north, then east. In this case, the easterly 
wave may become associated with a pressure trough and move 
northeastward as an extratropical cyclone. The most important char- 
acteristic of the easterly wave, however, is its tendency to incubate 
or propagate hurricanes and typhoons. 

2 Herbert Riehl, Tropical Meteorology. New York: McGraw-Hill Book Company, 
1954, pp. 237-238. 



1. From a series of weather maps: 

(a) Determine the size of the highs and lows within the closed 
isobars, noting the length and direction of the longest and 
shortest diameters, 

(b) Determine the direction and velocity of motion of the indi- 
vidual highs and lows. 

(c) Determine the distribution of temperature about the centers, 
Where is it highest? Where lowest? Are there sharp discon- 
tinuities of temperature? 

(d) Determine the distribution of cloudiness. 

(e) Where, with reference to the centers of the depressions, is rain 
falling at the time of observation? Where has it fallen during 
the past 24 hours? 

(f ) Determine the direction and velocity of the winds in the dif- 
ferent quarters of the lows and highs, and note their relation 
to the pressure gradient. Locate any definite wind-shift lines. 

(g) Check the location of the fronts by a consideration of wind 
directions, temperature differences, clouds, and rain. 

(h) Identify the depressions as young or as occluded lows. 

2. If a series of Washington weather maps showing West Indian hur- 
ricanes is available, make a similar study of these storms. 

3. Draw vertical cross sections of the frontal surfaces at several loca- 
tions across an occluded-wave cyclone. Can one tell by the cross-sectional 
sketch whether or not it is a cold-front-type occlusion? 




There are several air disturbances which, though smaller or less 
frequent than the cyclones and anticyclones of middle latitudes, are 
attended by characteristic and striking phenomena of major impor- 
tance wherever they occur. The largest of these is the tropical 
cyclone, seasonally familiar to certain ocean areas and coastal loca- 
tions in the low latitudes. More common to the interior locations of 
the United States are the thunderstorm and the tornado. In addi- 
tion, certain winds which are not independent disturbances but are 
parts of larger air movements acquire distinguishing properties in 
some regions and at times have received special names. Those to be 
discussed briefly in this chapter are the foehn, or chinook, sirocco, 
blizzard, and dust stonn. Special characteristics are acquired by 
these winds because of geographic situation, local topography, or 
conditions of the earth's surface. 

Tropical Cyclones 

A study of weather maps of the West Indies in summer and au- 
tumn shows an occasional low-pressure area, differing in a number 
of ways from the barometric depressions of higher latitudes, and 
traveling westward instead of eastward. Similar storms occur in low 
latitudes in other parts of the world, and the general name for them 
is tropical cyclone. Local names are hurricane in the West Indies, 
typhoon in the general western Pacific area, baguio in the South 
China Sea, and cyclone in the Indian Ocean. 



Characteristics. A tropical cyclone is a true revolving storm, 01 
vortex, a vast cyclonic whirl with a calm central core, or eye, result- 
ing from the rapidity of the whirling motion. There are no fronts 
separating masses of warm and cold air; temperature, pressure, 
winds, and cloudiness are more or less symmetrical around the 
center. Pressure gradients are steep, and winds often reach destruc- 
tive velocities of from 75 to 200 miles per hour (120 to 320 km per 
hour). For the storm to be classed as a true hurricane, the wind 
must reach a velocity of at least 75 miles per hour (120 km per 
hour). Winds are directed counterclockwise in the Northern Hemi- 
sphere and clockwise in the Southern Hemisphere, 

In well-developed hurricanes, the pressure at the center is below 
28.50 inches (965 mb) -often, much below. In 1933, four storms 
with minimum pressure between 27.40 and 27.99 inches (928 and 
948 mb) occurred in the region of the West Indies. A tropical cy- 
clone is an area of active convection, the air moving upward in 
spirals around the core. The closed cyclonic circulation usually be- 
gins at heights of one or two miles and builds down to (he surface 
and up to heights of from four to seven miles. Its diameter is usu- 
ally from 300 to 600 miles (480 to 960 km). Details of the structure 
of tropical cyclones have recently been obtained by airplane flights 
in and around them at various levels (Fig. 98), and by the use of 
radar, with raindrops echoing the microwaves. The Army radar sta- 
tion near Orlando, Florida, obtained an excellent reroul of the 
Florida hurricane of September 15, 1945. It found that the eye of 
the storm was 12 miles (19 km) in diameter, that the dense clouds 
of the whirling storm extended to an average height of 18,(X)0 feet 
(5.5 km), and that long "tails" or rain-bearing clouds spiraled around 
the storm center. 

Fig. 98. Panorama inside the Eye of a Typhoon. Photographs by Paul A. Humphrey 
on August 7, 1945, in the South China Sea. U. S. Navy Photo. 


As such a storm approaches, the barometer begins falling, slowly 
at first and then more and more rapidly, while the wind increases 
from a gentle breeze to hurricane force, and the clouds thicken from 
cirrus and cirrostratus to dense cumulonimbus, attended by thunder 
and lightning and excessive rain. These conditions continue for sev- 
eral hours, spreading destruction in their course. Then suddenly 
the eye of the storm arrives, the wind and the rain cease, the sky 
clears, or partly so, and the pressure no longer falls but remains at 
its lowest. This phase may last 30 minutes or longer, and then the 
storm begins again in all its severity, as before, except that the wind 
is from the opposite direction and the pressure is rising rapidly. As 
this continues, the wind gradually decreases in violence until the 
tempest is passed and the tropical oceans resume their normal re- 
pose. The violent portion of the storm may last from 12 to 24 hours. 

The Florida Keys storm of September 1935. A hurricane of great 
intensity devastated some of the Florida Keys on the afternoon and 
night of September 2, 1935. The center passed over Long Key, 
where a co-operative observer of the Weather Bureau, J. E. Duane, 
and 19 other persons were living at a fishing camp. The following 
are paraphrased from Mr. Duane's graphic and complete descrip- 
tion of the storm: 

September 2: 2 P.M. Barometer falling; heavy sea swell and high tide; 
heavy rain squalls continue; wind from N or NNE, force 6. 

3 P.M. Ocean swells changed; large waves now rolling in from SE, 
somewhat against winds, which are still in N or NE. 

4 p.M.-Wind still N, force 9; barometer dropping 0.01 inch every five 
minutes; rain continues. 

5 P.M. Wind N, hurricane force; swells from SE. 

6 p.M.-Barometer 28.04, still falling; heavy rains; wind still N, hurri- 
cane force and increasing; water rising on N side of island. 

6:45 P.M. Barometer 27.90; wind backing to NW, increasing; heavy 
timbers flying; beam 6 by 8 inches, 18 feet long, blown through ob- 
server's house. 

7 P.M. Now in main lodge building, which is shaking with every 
blast and being wrecked by flying timbers; water piling up on north side 
of camp. 

9 p.M.-No signs of storm letting up; barometer still falling very fast. 

9:20 p.M.-Barometer 27.22; wind abated. During this lull all hands 
gather in the last cottage. Sky is clear to northward, stars shining brightly 
and a verv light breeze continues; no flat calm. About the middle of the 
lull, which lasted 55 minutes, the sea began to rise very fast from ocean 


side of camp. Water lifted the cottage from its foundations and it floated. 

10:10 P.M. Barometer 27.02; wind beginning to blow from SSW, 

10:15 P.M. First blast from SSW, full force. House is now breaking 
up; wind seems stronger than at any time during storm. Barometer reads 
26.98 inches. I was blown outside into sea; got hung up in broken fronds 
of coconut tree and hung on for dear life; was then struck by some object 
and knocked unconscious. 

September 3: 2:25 A.M. Became conscious in tree and found I was 
lodged about 20 feet above the ground. The cottage had been blown 
back on the island, from whence the sea had receded and left it with all 
people safe. 

Hurricane winds continued till 5 A.M. and terrific lightning flashes were 
seen. After 5 A.M. strong gales continued throughout the day with very 
heavy rain. 

It is estimated that in this storm, wind velocities were 150 to 200 miles 
per hour. Destruction was practically complete over a path 30 miles 
wide, extending considerably farther to the right than to the left of the 
path of the center. The destructive storm tide had the same direction of 
advance as the storm center, flowing from southeast to northwest. The 
rate of advance of the storm was about 10 miles per hour, and the calm 
center was perhaps 8 miles in diameter/' 1 

Lowest observed pressures. It is not known how low the pres- 
sure may fall at the centers of severe hurricanes, because in the 
greater number of such storms no records are obtained. The lowest 
barometer readings of which there are reliable records are given by 
McDonald as follows: 

In the Florida Keys storm just described, an aneroid barometer in a 
boat tied up near the north end of Long Key indicated a barometric pres- 
sure of 892.2 mb (26.35 inches) at the center of the storm. This pressure 
was arrived at after careful tests of the aneroid with standard mercurial 
instruments under reduced pressure in the laboratory. It is the lowest sea 
level pressure ever observed in the Western Hemisphere. The lowest pre- 
vious record was 914.7 mb in the Caribbean hurricane of November 5, 
1932, and the previous record in the United States was 929.6 mb, Sep- 
tember 16, 1928, at West Palm Beach, Florida. Only one reading lower 
than the Florida Keys storm has been reported at sea level anywhere 
in the world. This was a pressure of 886.8 mb (26.185 inches), observed 
in a typhoon about 460 miles east of Luzon on August 18, 1927. 4 

3 W. F. McDonald, "The Hurricane of August 31 to September 0, 1935," Monthly 
Weather Review, Vol. 63 (1935), pp. 269-271. 

4 W. F. McDonald, "Lowest Barometer Reading in the Florida Keys Storm of Sep- 
tember 2, 1935," Monthly Weather Review, Vol. 63 ( 1935), p. 295. 



Regions and times of occurrence. Tropical cyclones begin over 
the oceans in the equatorial trough when it is some distance from 
the equator. They are more frequent on the western sides of the 
oceans, but some originate toward the eastern boundaries, as, for 
example, near the Cape Verde Islands in the north Atlantic and off 
the coast of Mexico in the Pacific. The six general regions of the 
world where most tropical cyclones occur are: (1) from the Baha- 
mas to the Caribbean Sea and the Gulf of Mexico; (2) in the Pacific 
Ocean west of Mexico and Central America; (3) in the neighbor- 
hood of the Philippines and the China Sea; (4) in the Bay of Bengal 
and, less frequently, the Arabian Sea; (5) in the southern Indian 
Ocean east of Madagascar; (6) in the south Pacific from the vicinity 
of Samoa and the Fiji Islands westward to the north and west coasts 
of Australia. Some tropical cyclones occur outside of these regions. 
None are known to occur in the south Atlantic Ocean. The regions 
of frequent occurrence and the normal paths are shown in Fig. 99. 

These disturbances occur almost exclusively in summer and au- 
tumn, in contrast to the depressions occurring within the prevailing 
westerlies, The latter are present at all seasons of the year but are 
most active in winter. In the Northern Hemisphere, tropical cy- 

Fig. 99. Regions and Generalized Patlis of Tropical Cyclones. Individ- 
ual paths are extremely variable, and occasional storms of this type develop 
or travel far outside of these* areas. 

clones occur from May to November, but the months of greatest 
frequency are September and October, except in the Arabian Sea. 
There they are most frequent in the calm seasons between the 
monsoons, that is, in June and again in October. The average num- 
ber in the North Atlantic is about six per year, but not all of these 
reach destructive violence. In the Southern Hemisphere, the time 
of greatest frequency is January to March, inclusive. 


Origin and path. Tropical cyclones originate in the warm, moist 
air of the equatorial trough. The winds are light and usually drift- 
ing lazily from east to west. A wave appears in the easterly flow and 
proceeds westward at from 10 to 15 miles (15 to 25 km) per hour. 
Why some easterly waves develop into raging tropical cyclones and 
others remain a relatively stable disturbance is not fully understood. 
They seldom have been known to form nearer than 5 nor more than 
20 from the equator. Some believe the Coriolis force is a necessary 
contributing factor to formation; hence no storms would form im- 
mediately at the equator, because there the Coriolis force is xero, 
Riehl and others have shown that tropical cyclones are frequent 
only over very warm oceans and are extremely rare when the sea- 
surface temperatures are below 79 F. (20"C.), 5 There are no storms 
in the South Atlantic nor in the eastern part of the South Pacific, 
where cold ocean currents keep the surface water temperatures 
continually below 79F. Nor do storms form in the regular source 
regions when the water temperatures are much below normal. 

Tropical cyclones, once started, move westward in the prevail- 
ing westward drift of the easterlies and curve gradually to the right 
in the Northern Hemisphere and to the left in the Southern. They 
travel at the moderate speed of from 10 to 30 miles ( 15 to 50 km) 
per hour, but occasionally remain stationary for a time or veer from 
their normal course. Finally, if they persist long enough, they move 
into the prevailing westerlies, often curving around the western 
sides of the summer oceanic highs, and then travel north-eastward 
with decreasing energy, becoming like ordinary extratropical lows, 
The typical path is a parabola, but the actual path of any given 
storm appears to be governed by the winds existing above it at the 
time. Records obtained in the region of the Caribbean Sea and the 
western Atlantic Ocean have shown that mature hurricanes move 
with nearly the same direction and speed as the wind directly above 
the closed cyclonic circulation. Twice in recent years, in September, 
1944, and again in September, 1954, hurricanes have entered the 
New England states, which are far out of their usual tracks* 

A tropical cyclone moving over land soon becomes larger and 
weaker and ceases to be of destructive energy. Evidently, this is 
because of the diminished supply of warm, moisture-laden air, and 

5 Herbert RicW, Tropical Meteorology. N'ew York; McGraw-Hill B<x>k Company, 
1954, p. 331. 


because of the increased friction over land. There are many storms 
over the oceans in the usual tracks of tropical cyclones that resemble 
hurricanes in many respects except that they fail to reach hurri- 
cane intensity. These have varying intensities from near-hurricane 
to moderate wind velocities. All are attended by rain. These facts 
indicate that favorable convective conditions must exist to a con- 
siderable height above the surface for the formation of a true hur- 
ricane. There are also many minor tropical disturbances, weak and 
poorly developed cyclones, over both land and water areas in the 
tropics, attended by squalls and thunderstorms. 

Effects of tropical cyclones. Tropical cyclones are destructive in 
their violence and are avoided if possible by ships at sea (Fig 100). 
The islands of the West Indies have been struck by hurricanes 
at various times, and paths of differing widths up to a few hundred 
miles have been laid waste, often with great loss of life. Storms 
of equal violence, killing large numbers of people, have also oc- 
curred in China, the Philippines, and Samoa. Occasionally storms of 
destructive severity reach Florida and the Gulf coast of the United 
States. There was such a storm at Galveston, Texas, in September, 
1900, with a loss of 6,000 lives, and in Florida in September, 1928, 
resulting in about 2,000 deaths. In the Florida storm, a large part of 
the loss of life and property was caused by the overflowing of Lake 
Okeechobee; strong north winds, estimated at 150 miles (244 km) 
per hour, raised the water level on the south shore by 10 to 15 feet 
and drowned many people. At Galveston, also, the loss of life was 
largely due to flooding of the low lands on which the city is built. 
( The city is now protected by a sea wall. ) 

The violence of these storms creates great ocean swells, which 
out-travel the storm and precede it by some distance. Along the 
Gulf coast, the water begins to rise when the hurricane is from 300 
to 500 miles (500 to 800 km) distant, that is, one or two days before 
the storm arrives, and often rises from 8 to 15 feet above the normal 
level of the Gulf. Observations of the direction and character of the 
waves as they reach the coast, and of the amount of the rise at dif- 
ferent places, afford a basis of forecasting the time and point at 
which the hurricane will arrive. 

The United States weather services follow as closely as possible 
the development and path of each West Indies hurricane and fore- 
cast its future movement and severity. They make use of reports 


from merchant ships moving through the region in various direc- 
tions, and supplement them by reports from land stations and recon- 
naissance aircraft. With the development of radiosonde and rawin 
soundings and their extension to great heights, upper-air reports 
have become of primary importance. From them, two methods of 
estimating with considerable accuracy the future movement of trop- 
ical cyclones have been developed. First, the direction and force 
of the wind at the steering level, that is, at the top of the closed 
circulation, give a reliable indication of future movement. The ob- 
jection to the use of this method is that the height of the steering 
level varies greatly in different storms and in the same storm from 
day to day. This fact makes it difficult to determine winds at the 
steering level without numerous soundings from many heights. 

In the second method, Simpson has shown that a tongue of 
wanner, lighter air develops in connection with a moving tropical 
cyclone and extends from 800 to 1,200 miles (1,300 to 2,000 km) in 
advance of the storm. 8 He finds that in all well-developed disturb- 
ances the warm tongue can be identified in the air layer between 
10,000 and 20,000 feet (3,000 and 6,000 m), and that its orientation 
in this layer gives a reliable indication of the path of the storm 
during the next 24 hours. 

During World War II, ships traveled in convoys, and ship radio 
reports ceased for security reasons, leaving vast areas of the hurri- 
cane region unreported. The United States Navy developed three 
methods of meeting this situation: (1) Aircraft reconnaissance 
methods were used to search out the storms. (2) Radar was uti- 
lized to locate the storm centers, as explained later. (3) Special 
instruments were designed to determine the direction and amplitude 
of microseisms originating at the storm center. Microseisms are 
feeble earth tremors detected only by specially constructed ap- 
paratus. Some of these tremors originate at the centers of intense 
lows and move outward in all directions with decreasing ampli- 
tude. By observing the direction and amplitude of these earth trem- 
blings at two or three coastal stations, it is possible to follow the 
path of a hurricane and estimate its intensity, and thereby to fore- 
cast its future movement and destructive force with some accuracy. 
None of the forecast methods have proved entirely satisfactory. 

* R. H. Simpson, "On the Movement of Tropical Cyclones," Transactions, American 
Geophysical Union, VoL 27, No. V, 1946. 



Broadly speaking, a thunderstorm is any storm in which thunder 
is heard. Thunder often occurs in tropical cyclones, general cyclonic 
storms, and tornadoes, but a typical thunderstorm, as distinguished 
from these storms in which thunder is incidental, is a local storm 
of short duration and of convective origin, proceeding from a large, 
anvil-shaped cumulonimbus cloud, often attended by heavy rain 
for short periods and sometimes by hail. 

Description of a local thunderstorm. On a quiet summer after- 
noon with gentle southern winds, a cumulonimbus cloud some- 
times approaches from the west or southwest, drifting east or north- 
east with the wind aloft while the surface air is moving slowly 
toward the cloud. The black, suspicious, threatening cloud draws 
near, and "Heaven's artillery thunders in the skies." About the 
time the first rain reaches the earth, there is a sudden strong and 
chilly gust of wind directly out of the storm and preceding it by 
several thousand feet. (Fig 101.) This out-blow may continue 
strong until the rain reaches the observer, then diminish quickly. 
The rain comes down in "sheets" for a time; then it also gradually 
diminishes, and in half an hour or so the storm is past, the sky 
clears, and a gentle wind again blows from the south. Such a storm 
is normally only a few miles wide, sometimes spreading over 30 
or 40 miles if it continues over a path 100 miles or more in length, 
as occasionally happens. The edges of the storm are well marked; 
the rainfall may be heavy within the path and diminish to nothing 
within a few hundred feet. 

Violent movements in a thunderstorm. If one watches for a time 
the growth of cumulus clouds with their flat bases and irregular, 
towering summits, one sees evidence of much turbulence and ac- 
tive vertical motion. Pilots are vividly aware of the dangers lurk- 
ing within these billowing clouds, especially after a first-hand en- 
counter with one. 

Because the cumulonimbus cloud is always a hazard to aviation, 
the United States Air Force, in co-operation with the Navy, the 
National Advisory Committee on Aeronautics, and the Weather By- 
reau, conducted a research project in 1946-1947 to determine more 


of the thunderstorm's characteristics. 1 In Florida during the summer 
of 1946, more than 500 penetrations of cumulonimbus clouds were 
made by skilled pilots at levels ranging from 5,000 to 25,000 feet. 

Fig. 101. A Mature Thunderstorm Showing Vertical Drafts and Area of Precipitation. 
Drawing by Bill Salwaechter. 

More than 800 similar penetrations were made of thunderstorms of 
Ohio the following jear. 

Violent updrafts and downdrafts exist side by side in the mature 
thunderstorms. The updrafts apparently reach greater velocities, up 
to 100 feet per second or about 70 miles per hour. Severe turbulence, 
including short, choppy gusts, together with the more steady ver- 

1 H. R. Byers and R. R. Brahani, Jr., The Thunderstorm. Washington, D. C.: U. S. 
Department of Commerce, 1949. 


tical drafts, create flying conditions so hazardous to aircraft that 
the storms should always be avoided whenever possible. 

Thunderstorm structure. Thunderstorms are made up of cells of 
circulation, each having its own vertical drafts operating independ- 
ently of the other cells. They are joined by "connective tissue" of 
static clouds. When thunderstorms persist over long periods of time, 
it is probable that new cells are forming and developing as old cells 

Three stages are recognized in the life cycle of a thunderstorm 
cell. First, the cumulus stage represents the early period of develop- 
ment when the entire cell is a single updraft current. The cloud is 
building vertically at a rapid pace, but no precipitation is possible. 
Occurrence of precipitation at the ground marks the beginning of 
the mature stage. Downdrafts, probably started by the drag of hy- 
drometers within the cloud, develop first in the lower portion and 
build upward through the cell. Throughout the mature stage, up- 
drafts and downdrafts persist in close proximity. Maximum intensi- 
ties of all aspects of the storm may be expected during this stage. 
Vertical development always extends well above the freezing level, 
and in some cases to a height of 65,000 to 70,000 feet (20,000 to 
25,000 m ) above sea level. The dispersal stage begins as the down- 
draft spreads over the entire cell. With the updraft cut off, the cell 
is no longer fed additional water vapor. The precipitation neces- 
sarily diminishes and then stops altogether. Much of the cloud struc- 
ture is shortly evaporated owing to the increasing mixing ratio re- 
sulting from the downdrafts. 

Individual thunderstorm cells may range in diameter from half a 
mile to 5 or 6 miles (1-10 km); however, measurements by the 
project sited above showed the average cell to be about 5,000 feet 
(1.8 km) in diameter. 

To produce the strong convectional activity necessary to the de- 
velopment of a thunderstorm, both an adequate supply of moisture 
and a large lapse rate are necessary. In order that the clouds may 
grow to sufficient height to produce a thunderstorm, an unstable 
condition must be created through a vertical distance of from two 
to five miles. This requires a lapse rate greater than the dry adia- 
batic to the lifting condensation level and greater than the wet 
adiabatic for a considerable distance beyond the freezing level 

Development of convective instability. There are several ways 


by which convective instability may be brought about. First, heating 
of surface air, such as occurs over land areas in summer, may create a 
large temperature difference between the lower air and the air above 
it. If the air is moist and conditionally unstable aloft, this gives rise 
to the typical thunderstorm described above and often called a heat 
thunderstorm. Such storms occur most frequently over land and on 
summer afternoons when the humidity is high. Although the air is 
cooled during the time of cloudiness and rainfall, it again becomes 
hot and oppressive after the storm has passed, for such storms occur 
within warm air masses. What may be called artificial heat thunder- 
storms sometimes occur over forest fires and active volcanoes, but 
only if the lapse rate above them is favorable. 

Second, the presence of abnormally cold air aloft, aided by con- 
vergence, may produce the necessary steep temperature gradient 
and instability. Such thunderstorms occur especially in the south- 
ern quadrants of depressions, where there are converging warm 
surface currents from the south or southeast and much colder upper 
currents from the southwest or west. They may occur by night and 
in winter, but over continental areas they are more frequent in sum- 
mer and by day, when local surface heating helps to create the nec- 
essary temperature contrast and when, also, absolute humidity is 

Over the oceans, convective thunderstorms occur mostly in win- 
ter and in the latter half of the night. There is little heating of the 
ocean surface by day not enough to produce strong convection 
currents. At night the ocean surface and the moist lower air cool 
slowly, while the upper air cools more rapidly by radiation. The 
difference in temperature becomes greater as the night progresses, 
and hence the lapse rates necessary for convection are most fre- 
quent late at night. Similarly, in winter the lower air over the oceans 
is relatively warm, because the water cools slowly, while the upper 
air is cold. 

The forcing of warm, moist air upward by its movement upslope 
or by the underrunning of cold air often furnishes the initial impulse 
in the formation of thunderstorms, when the lapse rate aloft is suf- 
ficient to continue active convection. Thunderstorms due to under- 
running cold air occur along an active cold front, sometimes in 
connection with general rains attending the passage of the front. 

Frontal and prefrontal thunderstorms may occur at any time of 


day and any season of the year but are rare over land areas in 
winter. Cold-front thunderstorms are followed by lower tempera- 
tures, because of the advancing cool air that causes them. When 
they follow a hot spell in summer, the newspaper headlines often 
say, "Showers bring cooler weather/' when the correct heading 
would be, "Cooler air brings showers/' Another popular error is the 
assumption that hail has caused the cooler weather. The cooling by 
hail is slight, temporary, and local. The change to cooler weather 
is due to the arrival of a cool air mass. 

Thunderstorm types. Thunderstorms are often classified in two 
main types, namely, air-mass and frontal Air-mass thunderstorms 
are those occurring as a result of vertical displacement of the air 
within a single air mass. The type includes local heat thunderstorms, 
induced by thermal convection, orographic thunderstorms due to 
movement of air against rising ground, and upper-level thunder- 
storms caused by advection of warm air at low levels or by over- 
running of cold air aloft. Frontal thunderstorms are the result of 
the interaction of two air masses in connection with the passage of 
a front. They are particularly characteristic of cold fronts. 

Geographic distribution. Thunderstorms are most frequent in 
the rainy regions of the tropics where heat and moisture are abun- 
dant and where, also, light winds favor convection. At some places 
within the tropics, as in Panama, Java, and equatorial Africa, the 
average number of days with thunderstorms is as great as 200 per 
year. They are rare in polar regions and in cold areas generally. In 
the United States they are most frequent along the eastern Gulf 
coast, where they occur on more than 70 days per year, mostly from 
June to September, inclusive, reaching an average of 94 a year at 
Tampa, Florida (Fig. 102). There is a secondary maximum for the 
United States in the southern Rocky Mountain region, Santa Fe, 
New Mexico, averaging 73 thunderstorm days per year. Here pro- 
graphic influences are the most important factor, because mountain- 
sides facing the wind force air upward, and mountainsides facing 
the sun are great aids to convection. The region of minimum fre- 
quency in the United States is in the Pacific coast states, where 
thunderstorms average from 1 to 4 a year, not including the moun- 
tain regions. 

It has been estimated that over the earth as a whole an average 
of 44,000 thunderstorms occur each day, and an average of 1,800 


Fig. 102. Average Annual Number of Days with Thunderstorms in the United States. 
From "Climate and Man" U.S.D.A. Yearbook, 1941. 

are in progress at all times. Because of their small size and local 
character, it is usually not possible to foresee the precise time and 
place of occurrence of thunderstorms, but their development and 
progress can be detected at a distance by the use of radar. The hour 
of fall and area covered by the rain seem to be matters of chance, 
especially for those storms in which local heating plays a large part. 
Thunderstorms occurring along a cold front can be placed more 
accurately, if the front is followed closely. 

Although thunderstorms over land areas are more likely to occur 
during the day than at night, because of the heating of the air by 
day, a large part of the United States receives more than half its 
precipitation at night during the warm season, April to September, 
inclusive. This is the region of the Great Plains, the Missouri Val- 
ley, and the upper Mississippi Valley. In this area of generally light 
rainfall, the occurrence of most of the precipitation at night is of 
some economic value in conserving the moisture. It is also of value 
in harvesting and threshing small grains and curing hay, because it 
permits drying by day. 


In this region of rather low average humidity, heating of the sur- 
face air is often not sufficient of itself to cause thunderstorms, but 
such storms are frequent when the lapse rate is increased by an in- 
flow of cold air aloft. This inflow occurs more often at night, ac- 
cording to Humphreys. Most of the summer thunderstorms of this 
region occur when there is a cool anticyclone along the northern 
border of the country between Montana and the Great Lakes, and 
when there is either low pressure in the southwest or a trough of 
low pressure across the central portion of the country from north 
to south, or from northeast to southwest. Under such conditions 
high daytime temperatures often prevail in the Great Plains and 
the Missouri and Upper Mississippi valleys. Hence, the lower air 
expands, and the pressure at an elevation of a half-mile and higher 
is increased until it may be approximately equal to that over the 
cold anticyclone to the north, at corresponding altitudes. This situ- 
ation prevents the inflow by day of much cold air at these heights, 
At night the warmer region normally loses heat more rapidly than 
the cooler region; the pressure at moderate heights accordingly 
tends to fall more over the warm region than over the cold, and 
this allows the cooler air to flow southward over the warm lower 
air. This result establishes that convectional instability is essential 
to the genesis of the thunderstorm. 

It is believed, also, that many nocturnal thunderstorms in this 
region are due to an increased inflow (advection) of warm air at 
night at altitudes between 3,000 and 6,000 feet (1 and 2 km). The 
reaction occurs between the warm air and the cold air above it, 
while the surface air cools by radiation and contracts. 

The electric charge. For two centuries since Benjamin Frank- 
lin made his famous "kite experiments," the mysteries of thunder- 
storm electricity have remained largely unsolved. It has become 
evident, however, that the cumulonimbus cloud is a huge static- 
electricity generator capable of building potentials of millions of 
volts within very short distances. The scientist, studying this phe- 
nomenon, is handicapped by being unable to create a working model 
of a thunderstorm in the laboratory, but ingenious instruments have 
been devised to measure the electrical characteristics within the 
actual thunderstorm. 2 

2 Ross Gunn, "Measurements of the Electricity Carried by Precipitation Particles," 
Thunderstorm Electricity. Chicago: University of Chicago Press, 1953, pp. 193-206. 


Experiments have shown that, when drops of water are broken 
into spray by a current of air, the spray particles gain a small posi- 
tive charge, and the drops that remain gain an equal negative 
charge. This may explain the positive charge observed in the lower 
front portion of the cloud, where rain is falling heavily through a 
rapid updraft of air. In the upper portion of the cloud, where tem- 
peratures are below freezing, it is thought that collision between ice 
particles causes the crystals to become negatively charged and the 
air positively charged. As the air ascends, it carries the positive 

Fig. 103. Direct Lightning Discharge. Courtesy, V. S. Weather Bureau. 


charge to the top of a cumulonimbus cloud. These suggestions ap- 
pear to offer probable partial explanations of the separation of elec- 
trical charges in a thundercloud. Other forms of cloud do not be- 
come thus highly charged, because of the absence of the rapid up- 
rush characteristic of cumlonimbus. 

Nature of lightning. Lightning is the flash of light caused by a 
discharge of atmospheric electricity. The discharge may be ( 1 ) be- 
tween two parts of the same cloud, (2) from one cloud to another, 
or (3) between a cloud and the earth. Thunder is the sound of the 
discharge, due to the sudden expansion of the air by heating. Air 
offers a high resistance to an electric current, and the passage of the 
current through it produces a rapid heating. Lightning is a direct, 
not alternating, discharge, and its duration is from 0.0002 second up 
to perhaps 1 second or more in a multiple discharge (successive 
flashes along the same path). 3 The current varies from a few thou- 
sand to 100,000 amperes, and the potential difference is of the order 
of 100,000,000 volts. 

The common names, forked, zigzag, and streak lightning, are used 
when the path of the discharge is visible, whether between cloud 
and earth or from one cloud to another (Fig, 103). The path of a 
discharge is never really a zigzag, but is often variously curved and 
frequently branching. Sheet lightning is the sudden lighting up of 
cloud and sky by a discharge the path of which is not seen. In this 
case the storm is usually distant, as indicated in Kipling's descrip- 
tion, "Sheet lightning was dancing on the horizon to a broken tune 
played by far-off thunder." Often the thunder is not audible. A rare 
and curious form of lightning, not fully explained, is known as ball 
lightning and consists of luminous balls or masses, usually moving 
at moderate speed, and lasting a few seconds (Fig. 104). The dis- 
turbing effects in radio receiving apparatus, known as atmospherics, 
sferics, or static, originate largely in lightning strokes, and thus the 
positions of large, distant thunderstorms can be determined by the 
use of two or more radio direction recorders (oscillographs) placed 
at known distances from each other. 

Protection against lightning. Lightning rods, if properly in- 
stalled, carry the electric current to the earth and afford good pro- 
tection to a building and its occupants. Proper installation requires 

3 J. H. Hagenguth, "The Lightning Discharge," Compendium of Meteorology. Bos- 
ton: American Meteorological Society, 1951, pp. 136-54. 


Fig. 104. B.tll I u'junim^ 'Ihirr stutfrs, about 2 l /2 minutes elapsed between 
tlu hist ami last stairs I'lmto by /, C. Jensen. 

that the conductors be of sufficient size, extend to every high point 
of the building, be cross-connected into one system with good joints 
and no sharp angles, and be well grounded at several places. Steel 
buildings arc sate places to be in during a thunderstorm, and any 
house is safer than out-of-doors, Low places are safer than hills. 
Wire fences and trees standing alone are especially to be avoided. 


A small storm which is rare* in its occurrence at any one place but 
which is much leaied because ot its destructive \iolencc is the tor- 
nado, meaning, in its deri\ation, a turning or whirling wind, and 
otteu colloquialK called a twister. 

Tornado characteristics. Tornadoes are revolving storms, turn- 
ing countoi clockwise in the Northern Hemisphere. They are of great 
intensity but small diameter and ha\e rapidly rising air at the cen- 
ter. They are haiometrie depressions resembling tropical cyclones 
but much smallei, of much shorter lite, and with much steeper pres- 
sure gradients. A funnel-shaped cloud de\elops in a low, heavy 
cumulonimbus cloud mass and extends toward the earth. The fun- 


nel rises and falls, turns and swings in various directions, Where 
it reaches the earth, there is almost total destruction attended by a 
deafening roar and by semidarkness; where it fails to reach the 
earth, there is little damage. It is estimated that winds near the 
center attain velocities of from 200 to 500 miles per hour ( 100-250 
mps), and the updraft at the center reaches very high velocities. 
The strongest natural winds that ever occur near the surface of the 
earth are associated with tornadoes, The funnel always develops in 
association with the lower portion of an exceptionally violent thun- 
derstorm. Heavy rain or hail may precede and follow the storm pas- 
sage, although some destructive tornadoes have been officially re- 
corded with no form of precipitation in the area. 

The pendant cloud develops downward from the base of the 
cumulonimbus (Figs. 105 and 106). It is a real cloud of water drop- 
lets formed by rapid expansional cooling of air entrained in (he cir- 
culation. Dust and other debris are pulled into the cloud as the 
funnel reaches the earth. The diameter of the destructive portion 
is generally less than a quarter of a mile, but paths of destruction 
range from a hundred yards to more than a mile in width. All fun- 
nels do not look the same. Many are clearly visible for several miles, 
as the one in Fig. 105; others are obscured by turbulent scud clouds 
extending down to the ground. 

TORNADO AT (.omi MU w., \IBHASKA, JUNK 24, 1930 

Fig. 105a. Heavy, Dark Cloud with Raided, Irregular Undersurface and 
Evidence that a Funnel Is About to Form. 


Fig. 1051). The Funnel Has Reached the Ground and Is Approaching the 


Fig. 105c. What Remains of Several Farm Buildings and an Automobile 
after the Tornado Has Passed, Photos by Otto Wiederanders. 



Fig. 106. Tornado at Rockwall, Texas, April 30, 1947. Courtesy U. S, Weather 
Bureau and Press Association, Inc. 

The speed of a tornado over the ground varies between different 
storms and with time during a single storm. The average cross- 
country movement, usually in a northeasterly direction, is at the rate 
of from 35 to 45 miles per hour (18-23 mps). While the average 
hourly speed of some tornadoes has been as low as five miles, at 
least one storm was clocked at 65 miles per hour over a distance of 
17 miles. Although, in rare cases, the funnel cloud has been reported 
to "stand still" for a few minutes, normally a tornado at a given 
place is all over in about 30 seconds. The path ranges in length from 
a few hundred feet to more than 100 miles. The length of path of 


more than a thousand tornadoes, however, averages about 10 or 
15 miles. The average tornado path covers an area of about three 
square miles. 

Destructive forces in a tornado. There are three damaging forces 
active in a tornado. First, the "hideous tempest" wrecks buildings 
and shakes down trees. Using the simple wind formula given on 
page 35, the pressure exerted against a vertical wall arranged nor- 
mal to the wind direction of a tornado would range from 160 to 
1,000 pounds per square foot. Second, there is an explosive effect 
within buildings because of the sudden reduction of pressure on the 
outside. Very few measurements have ever been recorded of pres- 
sures within the tornado funnel. A few observations have given the 
indication that sudden atmospheric-pressure drops, ranging from 
one to five inches of mercury, can be expected. This would leave an 
instantaneous net excess pressure within a tight building of from 
70 to 400 pounds per square foot. A building may literally "explode" 
if it is not constructed to withstand such internal pressures. Third, 
the lifting effect of the violent updraft may raise even heavy ob- 
jects and carry them considerable distances before dropping them 
to the earth, or sometimes set them down gently without damage. 

Place and time of occurrence. Every state in the United States 
has experienced one or more tornadoes. They are primarily an at- 
mospheric phenomenon of North America and Australia, and more 
especially of the central part of the United States, although they 
have occurred at irregular intervals on every continent of the globe. 
Conditions peculiarly favorable to tornado formation especially fre- 
quent those states east of the Rocky Mountains. 

The greatest tornado frequency per unit area during a recent 35- 
year period occurred in Iowa, which averaged 2.8 tornadoes per 
year per 10,000 square miles. Not far behind were Kansas, Arkansas, 
Oklahoma, and Mississippi. Texas has recorded more tornadoes than 
any other state because of its size, but the frequency per unit area 
is small. Tornadoes also occur rather frequently in Illinois, Indiana, 
Missouri, Nebraska, Alabama, Georgia, Ohio, Minnesota, Wisconsin, 
and southern Michigan. The number reported in the United States 
averages about 150 per year. The number recorded has been increas- 
ing recently, probably owing to an increasing population density 
and to improved methods of communication. There is no indica- 


tion that tornadoes are actually becoming more frequent. Some- 
times there is a concentration of storms in a given year or even on 
a given day in a single area. For example, 29 confirmed tornadoes 
occurred along a cold front across Oklahoma during the single after- 
noon of May 1, 1954. 

Most tornadoes occur during the spring months and during the 
afternoon hours of the day, but no month of the year or hour of the 
day has been completely free of storms. In fact, some of the more 
destructive storms have occurred out of season or at night when 
the people were caught completely off guard. Because of the small 
size and unusually short path of a tornado, the chances of a given 
building being wrecked by one are extremely small even in areas 
where these storms are most numerous. The same is true regarding 
the loss of life. 

Tornado warning. Forecasting tornadoes has been an extremely 
difficult task. The storm represents a local violent convection in the 
atmosphere which lasts only a short time. Even ioday, the exact 
requisites for tornado formation are not known in detail. One theory 
proposes that advancing cold, dry, polar air aloft flows ahead of its 
frictionally retarded surface front, thus extending itself in a very 
unstable condition over moist, warm air from the Gulf of Mexico. 
Finally, when a break-through occurs, to relieve the unstable situa- 
tion, the result is like pulling the plug in a filled bathtub. Another 
theory proposes that there is an upper cold front riding along the 
boundary surface of the warm air mass and more or less parallel 
to, and in advance of, a surface cold front. This involves the inter- 
action of three air masses, namely, mT, mP, and cP. The unstable 
conditions are then created by the upper, cold air mass advancing 
over warm moist air at lower levels. 

Several advances have been made in tornado-warning techniques 
since 1942. These are summarized by S. D. Flora and will be only 
briefly mentioned here. 4 The urgency of protecting military instal- 
lations caused the United States Air Force and the Weather Bureau 
to set up, in 1942, a fan-shaped reporting area to the west of sev- 
eral key bases. This service proved very satisfactory and was ex- 
panded after World War II. In addition, H. L. Jones, of Oklahoma 

4 S. D. Flora, Tornadoes of the United States. Norman, Okla.: University of Okla- 
homa Press, 1953, pp. 37-50. 


A. and M. College, developed a sf erics identification and tracking 
device which promises to be effective once the storms have formed. 5 
Morris Tepper, of the United States Weather Bureau, advanced a 
"pressure jump" theory in 1950, which has contributed to the un- 
derstanding of the dynamics of the tornado. 6 

Two unusual tornadoes on March 20 and March 25, 1948, struck 
Tinker Air Force Base near Oklahoma City and inflicted property 
damage of $10,000,000 and $6,000,000, respectively. As a result, the 
Air Weather Service set up a tornado research unit at the base, where 
Fawbush and Miller developed a workable method of forecasting 
tornadoes, 7 Modifications of this method are widely used for recog- 
nizing probable regions of occurrence. Criteria for the forecast con- 
sist of the recognition of an active cold front separating polar air 
from maritime tropical air in the same vicinity that a strong high- 
altitude jet of cold air from the west crosses a moisture tongue 
which is invading from the Gulf of Mexico or the Atlantic Ocean. 
Tornado-producing weather conditions can now be reliably identi- 
fied in advance of the storm and limits of possible tornado occur- 
rence determined. 8 The Weather Bureau inaugurated a public tor- 
nado-warning service in 1952. It consists of a "tornado alert" re- 
leased through radio and television mediums for a definite area dur- 
ing a specified time. 

Waterspouts. Where tornadoes occur at sea, they are known as 
waterspouts. The funnel cloud is formed in the same way, by un- 
stable atmospheric conditions, and is generally associated with a 
thunderstorm. When it reaches the water surface, it picks up spray. 
Such waterspouts are known to occur off the east coast of the United 
States, in the Gulf of Mexico, and off the coasts of China and Japan, 
in regions where cold, continental air extends over warm water. 
They have a cyclonic circulation like that of tornadoes (Fig. 107). 
Another type of storm that is also called a waterspout begins in fair 

5 H. L. Jones, A S/eric Method of Tornado Tracking and Identification, Stillwater, 
Okla,; Oklahoma A. & M. College, 1952. 

Morris Tepper, "On the Origin of Tornadoes," Bulletin of the American Me- 
teorological Society. Vol. 31 (Nov., 1950), pp. 311-314. 

7 E. J. Fawbush, R. C. Miller, and L. G. Starrett, "An Empirical Method of Fore- 
casting Tornado Development/* Bulletin of the American Meteorological Society, Vol. 
32 (Jan., 1951), pp. 1-9. 

* Lt. Col Ernest J, Fawbush and Major Robert C. Miller, "The Types of Airmasses 
in Which North American Tornadoes Form,*' Bulletin of the American Meteorological 
Society. Vol. 35 (April, 1954), pp. 154-105. 



weather instead of with a thunderstorm and is observed mostly in 
tropical waters. Such storms begin at the ground and grow upward, 
are small in diameter, are not much affected by the earth's deflective 
force, and may turn in either direction. There can hardly be strong 
enough contrasts of temperature in tropical waters to start these 
whirls, but it is thought that strong convection begins at the ground 
because the surface layer of air in contact with warm water be- 
comes very moist and hence lighter than the drier air above it. This 
convective rising is probably caused more by humidity differences 
than by temperature differences. 

Fig. 107. Small Waterspout. Photographed near Hong Korig along a cold front, 
August 8, 1945, by R. C. Fite. Note tin- line of wind-shift, as rvidrmvd on the sea 
surface. U.S. Navy Photo. 

Whirlwinds. Whirlwinds, or dust whirls, occur over land on hot 
days when the surface air becomes much warmer than that a few 
hundred feet above it, thus starting these small, shallow whirls of 
upflowing and inflowing air. By mixing the air to an increased depth, 
they prevent the surface air from getting as hot as it otherwise 
would. Unlike tornadoes, the whirls begin at the ground and may 
turn in either direction. They are common in many parts of the 
world, but especially in desert and semiarid regions, where they 
sometimes reach sufficient force to do some damage. 


Some Special Winds 

Many local popular names, in various parts of the world, have 
been given to winds coming from certain directions or having some 
easily recognizable characteristic. Often these winds have no general 
meteorological interest, but a few of them have special properties 
worth noting. 

Foehn or chinook. A foehn is a warm, dry, gusty wind of mod- 
erate to strong velocity, coming down a mountain slope. The move- 
ment is the result of pressure differences on opposite sides of the 
mountain chain. On the windward side, pressure is relatively high, 
and air is forced to rise over the mountain, with consequent expan- 
sion and cooling at the dry adiabatic rate, followed by condensa- 
tion and retarded cooling. On the leeward side, there is descent of 
air, compression, and adiabatic warming for the entire distance. 
Therefore, when the air reaches the same elevation at which it 
started on the other side, it has by the act of moving over the moun- 
tain become both warmer and drier ( Fig. 108 ) . Winds of this kind 
are local and intermittent in character. They are especially common 
on the northern side of the Alps in Switzerland, where they are 
called foehn winds, and on the eastern slope of the Rocky Moun- 
tains in Wyoming and Montana, where they are called chinooks. 
There is frequently a marked contrast between the air in these 
winds and the surrounding air, especially in winter, and the chinooks 
are capable of causing the rapid disappearance, by melting and 
evaporation, of a deep snow cover. With the arrival of a chinook, a 
great and rapid rise in temperature may occur. At Rapid City, South 
Dakota, on January 13, 1913, the temperature rose from 17 at 8 
A.M., to 47 at 10 P.M. 

Sirocco. Warm cyclonic winds have received local names in many 
parts of the world. A sirocco is a south wind coming from the Sahara 
Desert and reaching northern Africa hot, dry, and dusty. Sometimes 
it extends to the northern shores of the Mediterranean Sea in ad- 
vance of a low center moving eastward. Crossing the sea, it picks 
up enough moisture in the lower levels to become uncomfortably 
muggy. This condition may be intensified by the foehn effect on the 
lee shores of Sicily and Italy. Sirocco has sometimes been applied 
more generally to any hot> dry wind occurring in the warm sector 
of a moving depression which has been heated by blowing over a 



K>,000 FEET 

5000 FEET . 

10,000 FEET 

- 5000 FEET 


Fig. 108. Diagram of Foehn Wind and Resulting Changes in Air Condition. Con- 
densation begins at 5,000 feet, and the air cools at the wet adiabatic rate to the top 
of the mountain. Condensation and precipitation do not occur on the lee side, so the 
air warms at the dry adiabatic rate, returning to the base level much warmer than 

hot and arid land surface. Such hot winds occur over the Great 
Plains of the United States in the summer. 

Northers and blizzards. One of the most outstanding special 
winds of the central and southern parts of the United States, Mexico, 
and the Caribbean area is the norther. It is a strong, cold wind from 
a northerly direction in winter, caused by the rapid advance of a 
polar anticyclone. Its arrival is accompanied by rapid temperature 
falls, sometimes as much as from 20F. to 30F. in one hour, and at 
times by snow, sleet, or rain. Severe northers are sometimes referred 
to as cold waves. Occasional severe northers bring free/ing condi- 
tions and resulting damage to the truck-gardening and citrus indus- 
tries of the Gulf Coast and the Rio Grande Valley. A similar wind 
in South America is called the pampero. 

Blizzard is a term originating in America and refers to a violent 
cold wind (usually a norther) which is laden with snow, partially 
or entirely picked up from the ground. The snow usually consists 
of fine, powdery particles whipped by the wind in such great density 
that one can see only a few yards through it. The snow particles 
often are so fine and dense that they give the appearance of dense 
fog. Officially, the blizzard is defined as wind velocities of 32 miles 
per hour or more, accompanied by a temperature in the twenties 
or lower and much snow in the air, from either falling snow or from 
blowing snow originating on the ground, or both, reducing the visi- 
bility to less than 500 feet and occasionally to zero. Blizzards have 
been reported in Antarctica with velocities of from 75 to 100 miles 
per hour. 


Dust storms and dust falls. Moderate to strong winds blowing 
over a soil that is dry, loose, and unprotected by vegetation often 
raise clouds of dust which are carried along in the lower air by the 
wind. These are frequent in the southern Great Plains but are usu- 
ally local in their incidence. However, when great areas become ex- 
tremely dry, as happened in the early 1930's from North Dakota to 
Texas, and to a lesser degree in the early 1950's over the southern 
part of the same area, the entire lower atmosphere over large regions 
may be filled with dust. When the air has a stable lapse rate, the 
dust remains near the ground, and clear sky can be seen overhead. 
When the air mass is unstable, turbulence and convection lift the 
dust to greater heights, and a thick layer of the lower air becomes 
dust-laden so that the sky is overcast with a gray dust cloud and 
the sun becomes a pale disk or is completely hidden. Sometimes in 
the Great Plains region the cloud is so dense and dark in limited 
areas that artificial lighting is required at midday. Along a distinct 
front having a sharp increase in wind velocity, the cloud of dust 
may advance like a moving wall, and the very minute of its arrival 
at a given place may be observed (Fig. 109). At other times the 
dust diffuses and thickens slowly and imperceptibly. 

Dust which is thus lifted into the air is composed of fine particles 
which may be carried great distances, usually moving eastward be- 
fore settling to the earth. Thus, in the summer of 1934, dust originat- 
ing in the Great Plains was observed and collected in Washington, 
D. C. In the airway meteorological service of the Weather Bureau 
the condition is recorded as "dust" when dust is present and the 
visibility is from 1 to 6 miles, and "thick dust" when the visibility 
is less than 1 mile. In addition to producing disagreeable dust storms, 
turbulent winds over loose, bare soil cause much damage by drift- 
ing and by loss of fertile topsoil. Dust storms similar to those origi- 
nating in our Great Plains are frequent in the dry plains of northern 
China and in other parts of the world. 

When precipitation begins in an air mass carrying large amounts 
of dust, or falls through such a mass, the rain or snow gathers the 
particles as it falls and reaches the earth as "muddy" rain or discol- 
ored snow, leaving a coating of soil on exposed surfaces. Noticeable 
dust falls usually occur in this way, mixed with falling rain or snow. 
At Cheney, Nebraska, on May 12, 1934, there was a fall of colored 
hail, the stones looking much like small balls of clay, owing to the 


Fig. 109. Dust Storm, Johnson, Kansas, April 14, 1935. Courtesy, U. S. 
Weather Bureau. 

accumulation of yellowish dust. At Madison, Wisconsin, on March 
9, 1918, dust amounting to 13.5 tons per square mile was deposited 
with snow and sleet, giving it a light-yellow tint. Microscopic ex- 
amination of this dust showed that much the greater part of it was 
of mineral particles, ranging from 0.008 to 0.025 millimeter in size, 
hut it also contained fragments of leaves and other vegetable mat- 
ter, including fungi and spores. Its origin was probably in Oklahoma 
or Kansas. 

Occasionally a slackening of the dust-bearing currents permits a 
rapid settling of the dust, without accompanying condensation, re- 
sulting in a dry dust fall. At Lincoln, Nebraska, on April 29, 1933, 
there was a fall of dry, reddish-brown dust that continued for four 
hours, discoloring all horizontal objects and probably amounting 
to about 30 tons per square mile. Earlier in the day there were sev- 
eral showers of rain without any perceptible dust content, but four 
hours after the last rain, the dust began settling of its own weight, 
through quiet air, possibly aided by subsidence. 


1. Air starts at an elevation of 1,000 feet and a temperature of 65 and 
rises over a mountain at 7,000 feet, condensation beginning at 4,000 feet. 
What is the temperature of the air when it has descended on the other 


side to an elevation of 1,000 feet, assuming that the gain of heat by 
radiation and conduction is equal to the loss? 

2. If the air, before ascent in Problem 1, has a specific humidity of 7 
grams per kilogram and a mixing ratio of 14 grams per kilogram: 

(a) What is the relative humidity? 

(b) How much heat is imparted to the air for each gram of water 
precipitated if the specific humidity at the top of the moun- 
tain is 5 grams per kilogram? 

(c) If the mixing ratio is 16 grams per kilogram at 1,000 feet on 
the lee side, what is the relative humidity at that point? 

3. Why are thunderstorms very frequent at Tampa and at Santa Fe? 

4. Foehn and katabatic winds are both descending winds. Why is one 
warm and the other cold? 

5. Make a list of precautionary measures that would be valuable to 
remember in case an approaching tornado is sighted. 

6. The following storm advisories were issued by the Miami Weather 
Bureau Office regarding the location of the hurricane shown in Fig. 100; 

4:45 A.M., Sept. 19, 1948, 19.0N., 82.0W. 

4:45 A.M., Sept. 20, 1948, 20.7N., 82.4W. 

2:45 A.M., Sept 21, 1948, 23.6N., 8L6W. 

4:30 A.M., Sept. 22, 1948, 25.8N., 81.0W. 

5:00 A.M., Sept. 23, 1948, 28.7N., 78.7W. 

11.45 P.M., Sept. 23, 1948, 34.5N., 70.4W. 

11:00 P.M., Sept. 24, 1948, 42.6N., 56.2W. 

Plot these positions on a blank map and draw the path of the hurri- 
cane. Compute the speed of the hurricane for each time interval. 




A primary object of weather study is to understand why the 
weather is as it is today and what it will he like tomorrow. The 
hope of being able to foresee future weather conditions furnishes 
the principal incentive for maintaining meteorological services by 
the government, armed forces, and private organizations. The prac- 
tical value of accurate weather prediction is evident. 

The basic tool of a forecaster is the weather map. It makes pos- 
sible the visualization of weather conditions over large areas. The 
interaction of different air masses and the resulting development 
and motion of cyclones and anticyclones become evident. As one 
traces the weather from day to day on a series of weather maps, he 
will surely become aware that weather travels. This fact is of pri- 
mary importance in weather forecasting. Forecasting is a matter of 
charting atmospheric conditions and interpreting them in such a 
way as to be able to foresee the state of the weather in the future, 
usually for a short period of time. The aim in the following discus- 
sion is to indicate briefly the nature of the problems involved in 
forecasting the weather and the means thus far developed to solve 
these problems. No attempt is made to give a complete exposition 
of forecasting techniques. 

Analysis of Synoptic Charts 

From the beginning of the use of weather maps until recently, 
the chief attention of forecasters was centered upon the moving 
cyclones and anticyclones of the weather map, as revealed by the 
pattern of isobars. The chief effort was given to estimating the paths 



and velocities of these highs and lows and their influence on the 
weather as they passed. Since the forces controlling their movement 
and changing intensity were unknown in detail, rules for forecast- 
ing were necessarily empirical. After a long and intimate familiarity 
with weather maps, the forecaster developed the ability to judge 
with considerable accuracy what changes in existing weather to ex- 
pect under given conditions. With increasing daily exploration of 
the upper air, there is increasing knowledge of the physics of the 
air and an increasingly definite and scientific basis for forecasting. 

Estimating movement of lows and highs. The following precepts 
are some of those developed from experience with the synoptic 
chart of surface conditions before the role played by air masses and 
fronts was recognized. They are still valid but are now supple- 
mented by additional information, as will be noted later. In decid- 
ing where a cyclone or anticyclone will be at a later hour, 12 to 48 
hours in the future, consideration is first given to the normal or 
average direction and rate of movement of a disturbance in that 
position, and then to any reasons that may appear for a deviation 
from the normal. Usually the lows and highs move with the pre- 
vailing westerlies at an average velocity of about 30 to 35 miles per 
hour in winter and about 20 to 25 miles per hour in summer, and 
the normal movement always has an easterly component. There are 
large individual variations from the average, however, and each case 
must be considered separately. 

Some of the more general criteria used in forecasting the move- 
ment of a depression are: It tends to move with the same velocity 
and direction as during the past 12 to 24 hours, in the absence of 
other indications. Strong winds in front of a low retard it. A low 
tends to move parallel with the isobars in the warm sector, but to 
cross the isotherms, that is, to move toward an area of high tem- 
perature. It tends to travel toward the area where the greatest fall 
in pressure is occurring, as indicated by the barometric tendency, 
that is, by the amount and sign of the pressure change during the 
past three hours. Highs move toward the area where the greatest 
rise in pressure is occurring. 

As an aid in visualizing the pressure changes that are in progress, 
the three-hour pressure tendencies may be entered on the weather 
map, or on a separate map, and lines drawn connecting points of 
equal change. Such lines are called isattobars. A similar pressure- 


change chart may be drawn to show the changes in pressure that 
have occurred during the past 12 or 24 hours. Such charts make 
clearly visible the areas of rising and falling pressure and the amount 
of rise or fall, thus indicating the current direction and speed of 
movement of the pressure systems. They are regularly used by fore- 
casters in estimating future changes for short periods in advance. 

The distribution of pressure around the area under consideration 
influences its movement; the tendency is to move toward a region 
of small gradient and away from a region of steep gradient. A strong 
high east of a low, especially if the high is increasing in intensity or 
is nearly stationary, will retard the low or deflect it to the right or 
left. Two lows close together tend to unite. Consideration must also 
be given to the question whether the low is increasing or decreasing 
in intensity or is likely to do so within the forecast period. A low 
with a marked pressure gradient but with weak winds around it 
will increase in intensity. There are other such precepts, but these 
will serve to illustrate the nature of the problems presented to the 

Estimating the resulting weather. Having decided where the 
highs and lows on the map will be on the following day and how 
they will change, the question remains of how the new distribution 
of pressure will affect the weather. What will be the direction and 
force of the wind? What changes in temperature will occur? Will 
there be cloudiness or rain? In answering these questions, the fol- 
lowing fundamental facts are kept in mind: (1) The wind has a 
direct relation to the pressure distribution. If the latter is correctly 
foreseen, the wind forecast should be correct both as to direction 
and approximate speed. (2) Temperature changes are largely con- 
trolled by the wind: "Every wind hath its weather" more espe- 
cially, its temperature. The questions to be answered in this con- 
nection are: What is the temperature of the air that is expected to 
arrive, and how will it be modified as it moves? How much cooler 
or warmer is it than is normal for the season, or than the air now 
in the area for which the forecast is being made? (3) Cloudiness 
and precipitation usually attend a moving depression. 

The occurrence of rain in any given pressure stiuation is related 
to the topography and slope of the region and its position relative 
to mountain chains and bodies of water, because these affect the 
two essential factors, namely, the amount of moisture in the air 


and the amount and rate of the upward movement of the air. The 
question of where the moisture comes from is not always answer- 
able, because of our lack of complete knowledge of the movements 
of the upper air. In most cases, however, the source is evident. 

Placing of fronts. Since the development and elaboration of the 
polar-front theory, beginning about 1920, major problems in the 
analysis of the weather situation at a given time have been the plac- 
ing of fronts in their proper positions on the weather map, the iden- 
fication of air masses, and the determination of their physical char- 
acteristics. In solving these problems, it is first necessary to be fa- 
miliar with the general characteristics of air masses and fronts, as 
outlined in Chapters 9 and 10. In the placing of fronts from surface 
data, the following considerations should also be kept in mind: 
Fronts persist from day to day, and the position of a front on any 
one map is a consistent development from its position on the pre- 
vious map. The change in wind direction at a front is usually well 
marked, especially along cold and occluded fronts. 

A well-developed front is usually marked by an area of clouds 
and precipitation more or less parallel to it. The types and areas of 
clouds and precipitation are significant in locating a front, as, for 
example, the succession of cloud forms from cirrus to stratus preced- 
ing a warm front, and the relatively narrow band of cumuliform 
clouds attending a cold front. The dew point usually shows an ab- 
rupt change on the passage of a front. The front lies in a pressure 
trough and the isobars make an abrupt change in direction at the 
front. In the warm sector of an active depression, the isobars are 
nearly straight lines and uniformly spaced. The character and mag- 
nitude of the pressure change in the past three hours are signifi- 
cantly different on the opposite sides of a front. 

As previously stated, the properties of an air mass change as it 
moves away from the influences that have given it its characteris- 
tics. The modification is greatest in the lower layers, where there 
is interchange of heat to and from the earth and much turbulent 
mixing of the air, and where evaporation and condensation alter 
the moisture content These influences so change the lower air that 
it is sometimes impossible to recognize its source by its surface 
properties, particularly by its temperature. Usually, however, there 
is a definite temperature discontinuity at a front, more pronounced 
in winter than in summer. 


Upper-Air Analysis 

The primary role of air masses in the control of the weather hav- 
ing been recognized, the necessity of frequent and well-distributed 
upper-air observations is evident. It is only by such observations 
that the characteristics, alterations, and movements of air masses 
can be known and their effects on the weather foreseen. Much of 
modern progress in scientific meteorology has been wholly depend- 
ent upon the addition of this third dimension to weather observa- 

The data for upper-air analysis are obtained from radiosonde re- 
cordings of pressure, temperature, and humidity at computed 
elevations, also from pilot balloon or rawin observations of wind 
direction and force at known elevations, and from aircraft recon- 
naissance reports. In the absence of wind observations, the geo- 
strophic and gradient winds are computed from the pressure gradi- 
ent. The data thus obtained are used to construct maps representing 
atmospheric conditions at various distances above sea level. It is 
found that, with increasing height, the surface irregularities and 
abrupt changes in pressure, temperature, and wind tend to disap- 
pear and the large-scale features of the circulation become more evi- 
dent and are characterized by more gradual changes. Moreover, the 
circulation in these large-scale features at the higher levels largely 
governs the movement and behavior of the disturbances in the lower 
levels. These facts indicate the fundamental importance of upper- 
air soundings and upper-air charts. 

Analysis of air masses. In the analysis of air masses, an effort is 
made to ascertain ( 1 ) the extent and physical properties of each air 
mass, (2) the relations of the different masses to each other, and (3) 
the location, structure, and movement of the fronts along which the 
different masses meet. Structure of an air mass includes such prop- 
erties as the temperature, humidity, and lapse rate, at different 
levels; the degree of stability or instability of the air; whether it is 
stratified or well mixed; the existence of inversions, and whether 
these are due to warm currents of a different air mass or to sub- 
sidence of the upper air. A knowledge of the structure along the 
front involves ascertaining the slope of the front, the difference in 
temperature between the two masses, and the extent of mixing and 
turbulence at their surface of contact. Air-mass analysis, then, con- 


sists of a detailed study of the structure of the air. For successful 
application to forecasting, frequent observations at the surface and 
also a network of observations aloft are required. 

Cross sections. A vertical cross section of the atmosphere may 
be prepared by plotting the data obtained from radiosonde flights 
at stations in an approximate line across the country (see Fig. 110). 
The values at the several elevations for each station are entered on 
a vertical line at that station, extending from a base line represent- 
ing the earth's surface up to heights of three or four miles. Isopleths 
are then drawn for temperature, potential temperature, mixing ratio, 
or other elements shown by the soundings. Several such sections 
may be constructed, using different directions. Such cross sections 
show clearly the condition and structure of the air at the time of the 
soundings and are of much value to the forecaster in interpreting 
what is occurring in the air. They are especially helpful in fixing 
the position and slope of fronts. 

Identifying air masses. That part of the air mass which is above 
surface influences is slower and more conservative in its changes 
than is the surface air. That is one reason why upper-air observa- 
tions are more valuable than surface observations in identifying air 
masses. But even in the upper air, temperature and relative humid- 
ity may change greatly within a short period. A general uplift or 
subsidence of an air mass is of not infrequent occurrence; it results 
in dynamic cooling or warming and a consequent increase or de- 
crease in the relative humidity. Hence, upper-air observations on 
successive days may give quite different temperatures and humidi- 
ties in the same air mass and mislead the forecaster into thinking 
that one mass has been replaced by another. 

The potential temperature remains the same with changing eleva- 
tion as long as there is no condensation or evaporation, and the 
equivalent potential temperature is unchanged even by condensa- 
tion or precipitation. The equivalent potential temperature is there- 
fore called a conservative property of the air. It changes very slowly. 
( See page 103. ) It is altered slightly by the evaporation of rain or 
fog and may be changed by the absorption or radiation of heat or by 
mixing with other air. These processes act slowly on large air masses 
above the surface layers. Accordingly, an air mass can be identified 
more definitely by its equivalent potential temperature than by its 
actual temperature or even its potential temperature. 



Similarly, specific humidity is a much more conservative property 
of an air mass than is relative humidity. The latter changes rapidly 
when the temperature changes, but temperature differences in them- 
selves have no effect on specific humidity. The only processes that 
alter the specific humidity are the actual addition or removal of 
water vapor. The mixing ratio is conservative in the same way as the 
specific humidity. Dew point is also a more conservative element 
than relative humidity, because dew point is a function of absolute 
humidity, which changes more slowly than relative humidity. But 
absolute humidity and dew point both change when pressure 

Another property, more conservative than dew point or specific 
humidity, is the wet-bulb potential temperature, which is the wet- 
bulb temperature that a parcel of air will have when brought adia- 
batically to the standard pressure of 1,000 millibars. The wet-bulb 
temperature is the lowest temperature to which air can be cooled 
by evaporation of water into it. Hence, although evaporation re- 
duces the temperature of the air, it leaves its wet-bulb temperature 
unchanged. In like manner, the condensation of moisture releases 
latent heat but does not alter the wet-bulb temperature. Accord- 
ingly, the wet-bulb potential temperature is not changed by adia- 
batic processes nor by the gain or loss of moisture. It varies only 
by acquiring heat from outside itself or by losing heat by conduc- 
tion, radiation, or mixing. Emphasis upon these conservative phys- 
ical properties has come with the development of upper-air obser- 
vations and air-mass analysis. They were little used when observa- 
tions were confined to the surface air. 

As was pointed out in Chapter 5, the basic tool for upper-air analy- 
sis is the adiabatic chart. From it can be read the degree of stabil- 
ity, mixing ratio, specific humidity, dew point, lifting condensation 
level, temperature, potential temperature, equivalent potential tem- 
perature, pressure, and height of all levels included in the sound- 
ing. These values are not only important in the analysis of atmos- 
pheric characteristics above a given station, but they may be trans- 
posed to several other specialized charts to provide regional, con- 
tinental, or even hemispheric coverage of the selected charac- 

The Rossby diagram (or equivalent potential temperature dia- 







I 2 3 4 5 6 7 6 9 10 II 12 13 14 15 
Mixing Ratio, Qromt pr Kilogram 

Fig. 111. Typical Air-mass Curves on Rossby Diagram. Sloping lines are lines of 
equivalent potential temperature. 

gram), developed by C.-G. Rossby, is one graph upon which the 
data obtained by upper-air soundings are plotted. Vertical distances 
(ordinates) on this diagram represent the potential temperatures of 
the dry air, and horizontal distances (abscissas) represent the mix- 
ing ratios. Equivalent potential temperatures are represented by 
sloping lines (Fig. 111). These are all conservative properties and 
are therefore useful in identifying vertical columns. The diagram 
facilitates the computation of the equivalent potential temperatures 
and displays graphically the characteristics of air masses. The dif- 
ferences between the different air masses are brought out clearly; 
each mass, from polar continental at one extreme to tropical mari- 
time at the other, has a characteristic curve and a characteristic 
position on the diagram. 

Constant-pressure (pressure contour) charts. The constant-pres- 
sure chart, also called the pressure contour chart, has become the 
most commonly used type of synoptic upper-air chart (Fig. 112). 
Fixed pressure levels above the surface are chosen, for example, 850, 


700, 500, or 300 millibars. In fact, a chart may be constructed for 
each of these levels. Height (in feet), 1 temperature (in degrees 
Centigrade), and dew point are obtained for each station at the 
chosen pressure level from the radiosonde reports. These data are 
entered on a chart at the geographical location of the respective 
stations. If rawin or pibal data are available for that level, wind di- 
rection and force may be plotted also. Isotherms are then drawn 
which show the distribution of temperature on this isobaric surface. 
Along such an isotherm, pressure, temperature, potential tempera- 
ture, and density are constant (neglecting the slight influence of 
moisture content on density ) . Contour lines are also drawn, connect- 
ing points of equal elevation. These lines show how the height at 
which the chosen pressure occurs differs from place to place. See 
Fig. 112, where contour lines are drawn for each 200 feet of differ- 
ence in the elevation of the 500-millibar pressure surface. 

Since they are all on the 500-millibar surface, each contour line 
is also an isobar. Thus the 18,000-foot contour on the constant- 
pressure surface is also a 500-millibar isobar on the 18,000-foot ele- 
vation surface. It follows, then, that the contour lines show the 
horizontal variation of pressure. For example, Fig. 112 shows an area 
inclosed by the contour of 17,600 feet. If we rose 400 feet above 
this area to a height of 18,000 feet, the pressure there would evi- 
dently be less than 500 millibars, for pressure decreases upward. 
Therefore, a low level of the contour lines indicates low pressure, 
considered horizontally, and a high level indicates high pressure. 

At this altitude, the highs and lows and the irregularities of the 
isobars, as found at the ground, give place to a smooth, wavelike 
succession of troughs of low pressure and ridges (or wedges) of 
high pressure. There is a mathematical relation between wave 
length and speed of advance, and a formula has been developed for 
calculating the movements of the series of troughs and ridges. There 
is also a theoretical relation between the speed and the relative po- 
sitions of isobars and isotherms in such wavelike motions. The move- 
ments of these troughs and ridges can thus be anticipated with fair 
accuracy for some days in advance. They are closely related to the 

1 In meteorology, height is usually determined by computation from the distribu- 
tion of temperature and pressure through the atmosphere. Heights are therefore ex- 
pressed in "geopotential" or "dynamic" units, which are almost the equivalent of 
linear units. 


movements of air masses and fronts at the earth's surface. In gen- 
eral, surface disturbances move along the isobars between the prin- 
cipal troughs and ridges shown on the upper-air chart and at a rate 
proportional to the pressure gradient between them. This chart is 
therefore of direct use to forecasters for daily forecasting and espe- 
cially in the preparation of forecasts for periods of two to five days 
in advance. 

Constant-pressure charts are usually drawn for several different 
pressure values. Pressures of 500, 700, and 850 millbars are most 
frequently used, corresponding, approximately, to altitudes of 
18,000, 10,000, and 5,000 feet, respectively. Contour intervals of 200 
feet are used on these charts. Charts are also in use for pressures of 
1,000 millibars (very near the surface) and for 300 millibars (about 
10 km or 30,000 ft. ), and sometimes for even smaller pressure values, 
extending into the stratosphere. 

On a pressure-contour chart, the geostrophic wind is a function 
only of the spacing between contour lines, that is, of change of 
height with change of horizontal distance (neglecting local differ- 
ences in gravity and in the Coriolis force ) , No correction for density 
is required. The direction and force of the wind at various pressure 
levels can thus be readily determined by the use of a single geo- 
strophic wind scale, the same scale for all pressure contour charts. 
The relative values of the wind at different levels are visible on the 
map. This is of direct value to the airplane pilot, and a knowledge 
of geostrophic wind values is important in forecasting the future 
movement of pressure systems. 

By comparing two charts at different pressure levels, the thick- 
ness of the layer between them can be determined. This thickness 
is a measure of the mean virtual temperature of the layer. (Virtual 
temperature is affected by the moisture content of the air and 
may be defined as the temperature of dry air having the same pres- 
sure and same density as the existing air with moisture present.) 
The warmer the layer, the greater the thickness. A further compari- 
son is obtained by superimposing successive surfaces one upon an- 
other. This procedure furnishes a check upon the accuracy and con- 
sistency of the analyses of the individual levels and brings out the 
relations between the various elements within the layer. Such an 
analysis of the atmosphere, layer by layer, is known as differential 
analysis. There is a shear of the geostrophic wind with height, that 



rrt ^ 






is, a change in direction and speed due to differences in temperature 
along isobaric surfaces. This is called the thermal wind component 
of the geostrophic wind aloft. It is determined in direction and mag- 
nitude by the distribution of the mean temperature. Hence, it may 
be obtained from the varying thickness of the layers between iso- 
baric surfaces. 

Knowing the winds and the mean virtual temperatures at various 
pressure levels in the free air helps the forecasters to estimate fu- 
ture pressure and temperature changes at the earth's surface. For 
example, it is found that there is a tendency for low-pressure centers 
to move along isotherms of mean virtual temperature between the 
1,000- and the 700-millibar surfaces at a speed proportional to the 
speed of the geostrophic wind. The temperature differences on an 
isobaric surface are obviously not due to differences of pressure. 
They are "real" and of real significance to the forecaster. Where the 
isotherms cut across the contour lines, there is an active movement 
of masses of warm and cold air, and a change of pressure is in 
progress. The direction of the thermal winds in relation to the con- 
tours has a "steering" effect upon these pressure changes; it indi- 
cates areas of pressure rise and pressure fall. 

Isentropic analysis. Another method of charting upper-air con- 
ditions is to select a surface having the same potential temperature 
throughout ( called an isentropic surface ) , and enter on the map at 
each station the pressure and the condensation pressure existing at 
the chosen potential temperature. Condensation pressure (also 
called saturation pressure) is the pressure at which saturation is 
attained in air ascending adiabatically. It can be readily obtained 
from the radiosonde observations. Lines may then be drawn ( isobars) 
connecting points of equal pressure, and other lines ( condensation 
isobars) connecting points of equal condensation pressure (Fig. 
113). Where corresponding isobars and condensation isobars are near 
each other, the air is moist; where they are far apart, the air is dry. 

The isentropic chart represents an attempt to study one of the 
most conservative properties of the atmosphere, namely, potential 
temperature. It is seldom used as a forecast tool, however, because 
similar information can be obtained from the constant-pressure 

Circulation index. As has been noted in Chapter 8, the prevail- 
ing westerly zonal circulation of middle latitudes results from the 


- OJ 



"1 1 





difference in pressure between the subtropical belts of high pres- 
sure and the subpolar low-pressure areas, and is modified by the 
cellular circulations around centers of action. These centers of ac- 
tion vary in intensity and position, not only with the seasons but 
also in short periods of irregular length, usually of a few weeks. The 
varying intensity of this zonal circulation in the prevailing westerlies 
of the Northern Hemisphere is expressed in terms of a circulation 
index. This index is obtained by taking the difference in sea-level 
pressure between latitudes 35N. and 55N. along several meridians 
around the hemisphere, and averaging these differences. If the aver- 
age difference is 8 millibars or more, the circulation is said to have 
a high index; if the difference is less than 3 millibars, it has a low 
index. The higher the index is, the stronger are the prevailing 

The index thus obtained is in terms of sea-level pressures, but the 
different index values are most easily and quickly recognized on 
the 700-millibar pressure contour chart, where they form distinctive 
patterns. When the index is high, the 700-millibar map shows nearly 
straight contours or long waves, and the troughs are small and move 
rapidly. At the same time, the centers of action in the lower air are 
large and few, the westerlies are strong, and there is little cyclonic 

When the index is low, the waves are short and move slowly, the 
centers of action break up into numerous small centers, the westerlies 
become weak, and there is much cyclonic activity. Briefly, rapidly 
moving waves on the upper-level charts indicate a high index, and 
slowly moving waves, a low index. When values of the index are 
intermediate, lying between about 3 millibars and 8 millibars, the 
conditions shown on the charts are also intermediate. A low-index 
pattern is indicated in Fig. 114, which shows marked cyclonic activ- 
ity and much north-south air movement. Severe blizzards and dust 
storms occurred in the Middle West concurrently with this synoptic 

Both high and low indexes occur characteristically in winter; sum- 
mer values are usually intermediate. A particular index may per- 
sist for from three to eight weeks and then give place to an index 
of the opposite type. Changes in the value of the index are evi- 
dence of variations in the zonal circulation; such variations largely 
determine the formation and movement of air masses and fronts and, 



hence, the general character of the weather. Therefore, the fore- 
caster must have in mind the existing index and its recent history. 
This is especially necessary in winter and in the preparation of fore- 
casts for several days in advance. 

Mean weather charts. Averaging values to obtain more repre- 
sentative results is not new, however, its application to weather 
analysis is fairly recent. Mean charts tend to eliminate lesser per- 
tubations in the atmosphere and to reveal only the larger or long- 
range trends. The technique has been used in climatology for many 
years. It promises to be of much value to meteorologists also, espe- 
cially for long-range forecasting, 

A five-day time mean chart, developed by Jerome Namias, 2 is 
drawn for the entire Northern Hemisphere on the basis of the five- 
day average height of a selected pressure surface (usually 500 milli- 
bars) over predetermined stations. This revealed a simplified pattern 
of slow-moving, long waves about the hemisphere which could defi- 
nitely be associated with regional weather characteristics. Since the 
most current five-day mean chart is necessarily 2V*> days old, how- 
ever, its value for operational forecasts is seriously impaired. The 
space mean chart was developed to overcome this handicap. It is 
obtained by averaging the values at the four points of a diamond 
and plotting the result at the center. 3 The space mean chart seems 
to possess all the merits of the time mean chart plus the fact that it 
can be analyzed and ready for forecasting use within a few hours 
of the actual observations. 

Forecasting the Weather 

The first process in the making of a weather forecast is the collec- 
tion of all available synoptic surface observations and recent upper- 
air soundings. The object is to obtain, as completely as possible, a 
picture of existing weather conditions over a large area. For fore- 
casting in the United States, the area should include the whole of 
North America, the North Pacific Ocean, and the western portion of 
the North Atlantic, in each case extending northward into polar 

2 Jerome Namias, Extended Forecasting by Mean Circulation Methods. Washing- 
ton, D. C.: U. S. Department of Commerce, 1947. 

a F. A. Berry, W. H. Haggard, and P. M. Wolff, Description of Contour Pattern* 
at 500 Millibars. Norfolk, Va.: Bureau of Aeronautics Project AROWA, U. S. Navy, 
April, 1854. 


areas. Reports are received from points in Alaska north of the Arctic 
Circle. The desired distribution of reports is not always fully at- 
tained. The establishment of a network of reporting stations in 
Arctic regions of Canada, Greenland, and Iceland has added to the 
value of the maps for forecasting purposes, and in particular has 
aided the establishment and maintenance of airway routes across 
Arctic regions. 

The primary purpose in collecting all this information is to be 
able to foresee future weather by understanding and interpreting, 
as far as possible, the physical conditions and physical causes of the 
existing weather and of future changes. A second important pur- 
pose is service to the public by furnishing information of existing 
weather conditions, and especially service to aviation by reporting 
conditions along all air routes. 

Charting of data. The following types of data are available for 
charting at the forecast centers : ( 1 ) the simultaneous surface ob- 
servations from which are prepared the synoptic weather map and 
supplementary charts, such as pressure-change charts; (2) radio- 
sonde observations, giving pressure, temperature, and humidity of 
the air at various levels above the stations; (3) pilot-balloon obser- 
vations and rawins, giving wind direction and force at known eleva- 
tions. These upper-air data are the basis for the construction of 
constant-pressure charts, winds-aloft charts, Rossby diagrams, isen- 
tropic charts, vertical cross sections, and mean pressure charts of 
the atmosphere. Not all of these facilities, but sometimes a few 
additional ones, are available for every forecaster. Making use of 
all the relevant information thus collected and charted, the fore- 
caster identifies the air masses and places the fronts properly on the 
weather map. Areas of precipitation and other special weather con- 
ditions are carefully analyzed and related to their causes. This is 
the modern method of analyzing the weather situation at a given 
time, and is known as weather analysis or air-tna$s analysis. 

Types of forecasts. Forecasts may be classified with reference 
to how far into the future the forecasters attempt to foresee weather 
conditions, such as, ( 1 ) daily forecasts for periods of from 12 to 48 
hours in the future; (2) short-range forecasts for from 1 to 12 hours 
in advance; (3) extended forecasts for 5- or 6-day periods ending a 
week after the time of issue; (4) long-range forecasts for the fol- 
lowing month, or season, or year. 


A forecast is a prediction of the occurrence of a condition, while 
a warning means that a dangerous condition has occurred or is in 
the state of occurring. Various special forecast and warning services 
are listed below: 

1. Ocean and Lake Forecast and Warning Service. Includes 
weather elements unfavorable to shipping. 

2. River and Flood Forecast and Warning Service. Includes run- 
off potential and flood stages of the nation's principal rivers. 

3. Hurricane Warning Service, centered at Miami, Florida. Col- 
lects and disseminates information on the location, direction of 
movement, and intensity of tropical cyclones. 

4. Horticultural and Agricultural Forecast Service. Includes 
forecasts and warnings of particular weather elements directly af- 
fecting the growing of crops and livestock. 

5. Fire-Weather Forecast and Warning Service. Designed to 
help prevent or combat forest fires. 

6. Cold Wave and Related Warnings. Serves cattlemen, ship- 
pers, and the general public when there is danger of the occurrence 
of frost, freezing temperatures, glaze, heavy snows, strong winds, or 

7. Severe Local Storm Warning Service. Is concerned with dan- 
gerous local storms, such as tornadoes, or thunderstorms accom- 
panied by severe lightning, heavy hail, or strong winds. 

8. Aviation Forecasts. Provides regional, terminal, and cross- 
country (trip) forecasts for aviation activities. 

In addition to the United States Weather Bureau, there are sev- 
eral other agencies providing several types of weather forecasts. 
Meteorologists of the air transport companies make many short- 
range forecasts along specific routes for the guidance of their own 
operations. Army, Air Force, and Navy aerologists, likewise, prepare 
forecasts for the use of their organizations and to meet their special 
needs. Since World War II, there has developed a nucleus of pri- 
vate weather forecasters in the United States. 4 Their services are in 
demand by individuals and companies whose operations may be ad- 
versely affected by unexpected changes in the weather. Each private 
forecaster, therefore, concentrates his efforts on forecasting the par- 
ticular variables of weather that are important to his employer. 

4 Joseph J. George, "On 'Weather is the Nation's Business/ " Bulletin of the Amer- 
ican Meteorological Society, Vol. 35 (Feb., 1954), pp. 43-47. 


Daily forecasts* Having charted the data, the next step in the 
preparation of the daily forecast is to estimate the movement, fu- 
ture position, and development of the patterns shown on the charts 
(Figs. 115, 116, and 117). The simplest procedure is to assume that 
the isobars, the fronts, and the areas of high and low pressure will 
move without change in the same direction and at the same speed 
as they are currently moving, as shown by the pressure-change 
chart. This procedure will usually give fairly accurate results for 
12 hours in advance, but beyond that period there is increasing 
error, and the forecaster must look for indications of change. Here, 
the circulation index and the upper-air charts, especially the pres- 
sure contour charts for 500 and 700 millibars, give valuable infor- 
mation. From these the forecaster draws prognostic charts for vari- 
ous constant-pressure levels, that is, he draws charts for these levels 
as he thinks they will appear at a selected hour in the future, usu- 
ally 24 hours in advance. These serve to check the indications shown 
on the surface weather map concerning the movements of pressure 
systems and fronts. 

Since fronts have a more or less definite association with centers 
of high and low pressure, the general rules applying to the move- 
ment of cyclones and anticyclones may be applied also with lit- 
tle modification to the movement of the discontinuities attending 
them, A few precepts relative to the movement of fronts may be 
summarized as follows: The movement of fronts, as well as of pres- 
sure systems, is more or less clearly indicated by the existing dif- 
ferences in pressure around them and by the rate of change of this 
pressure. More specifically, a warm front moves faster, the greater 
the fall in pressure in front of it within the preceding three hours, 
and a cold front moves faster, the greater the rise in pressure be- 
hind it. A front moves slowly when it is nearly parallel to the isobars 
and increases in velocity as the number of isobars intersecting it 
increases. Fronts are retarded by high mountain ranges and by large, 
slow-moving anticyclones. 

A method of computing numerically the movement of isobars and 
of pressure troughs and wedges on the surface map has been de- 
veloped by S. Petterssen. 5 The method is based on the pressure 
tendencies as shown on the surface map. If we take, for example, a 

5 S. Petterssen, Weather Analysis and Forecasting. New York: McGraw-Hill Book 
Co., 1940. 


station on the forward side of an advancing trough of low pressure, 
it is evident that the fall in pressure at that station during the past 
three hours is an indication of the speed of advance of the low, pro- 
vided the low is not changing in shape nor in intensity. Similarly, 
a station at the rear of the trough will show a rise in pressure due 
to the movement of the trough. If there has been no change in in- 
tensity, the fall in pressure at the station in front should equal the 
rise at a station an equal distance in the rear. This value, the same 
on each side, is therefore a measure of the speed of displacement 
of the trough. 

A change in intensity is indicated by a deepening of the low (pres- 
sure lower at the center than it was at the previous observation ) , or 
by a filling of the low (pressure becoming higher at the center). If 
there has been deepening, the station ahead of the front will show 
a pressure fall greater than the pressure rise at the rear. Similarly, 
if there has been filling, the advance station will indicate a smaller 
pressure fall than the pressure rise at the rear station. From these 
considerations, Petterssen developed a simple equation for comput- 
ing in miles per hour the movement of isobars and pressure sys- 
tems. His equation has proved valuable and is being used, although 
it does not take account of all possibilities of change and, therefore, 
sometimes gives erroneous results. 

A more recently developed technique for computing movement 
and intensity of surface weather features has been developed by 
J. J. George and associates/ 3 This method is now being widely tested 
by forecasters and has become known as the George method. It is 
an empirical manipulation of pressure, temperature, and wind 
values, taken from the synoptic weather chart and the 850-, 700-, 
and 500-millibar constant-pressure charts, to obtain a 24-hour fore- 
cast of movement and intensity of highs, lows, and fronts. 

Making use of all such prognostic precepts and rules, the daily 
forecaster next makes a prognostic chart of surface conditions, plac- 
ing isobars and fronts as he thinks they will appear at the end of the 
forecast period and perhaps, also, at intermediate intervals. He is 
then ready to prepare the detailed daily forecasts for a given sta- 
tion or a given area. The forecasts state by 12-hour periods the tem- 

Joseph J. George, et al, Forecasting Relationships Between Upper Level Flow 
and Surface Meteorological Processes. Cambridge, Mass.: Geophysics Research Di- 
rectorate, Air Force Cambridge Research Center, August, 1953. 


perature changes expected and whether or not precipitation is ex- 
pected, and often other details. 7 

Short-range forecasts. Ordinarily, the making of short-range 
forecasts consists in extrapolating for a few hours in advance the 
conditions shown on the current synoptic map. In doing this, atten- 
tion is given to the general trend of the weather situation and to any 
specific indications of change. For example, the Petterssen method 
of computing displacements is especially applicable to short-range 
forecasts. By applying these methods as they are used in making 
daily forecasts, short-range forecasts can usually be made without 
serious error and with greater accuracy in detail and in timing than 
is possible for longer periods. For this reason, air transport com- 
panies depend very largely on short-range forecasts in the operation 
of airports and the maintenance of flight schedules. In the case of 
the development of dangerous storms of small area, such as hurri- 
canes, tornadoes, or severe thunderstorms, these methods do not 
make it possible to foresee with the desired precision the severity, 
path, and time of arrival of such storms. In these cases, aircraft re- 
connaissance and radar storm detection have proved of great value. 

Radar storm detection. The development of radar and its appli- 
cation to weather phenomena marks a great advance in our ability 
to follow and foresee the movement of stormy conditions, and hence 
in the accuracy and precision of short-range forecasts, especially in 
connection with severe storms. Precipitation areas and clouds asso- 
ciated with precipitation reflect radar pulses back to the sending 
station and permit the forecaster to determine the position of such 
areas and to follow their movement minute by minute. The echoes 
returned by weather phenomena can usually be distinguished easily 
from those returned by other targets, and the different types of 
storm give distinctive echo patterns. In particular, the following 
weather conditions can be detected and distinguished from one an- 
other by radar methods, and their position and activity determined: 
tropical cyclones, tornadoes, thunderstorms, well-defined cold fronts 
and squall lines, general precipitation areas, and active convective 

7 For further details on the techniques of forecasting, see V. P. Starr, Basic Prin- 
ciples of Weather Forecasting. New York: Harper & Bros., 1942, and Berry, Bolay, 
and Beers, Handbook of Meteorology. New York: McGraw-Hill Book Co., 1945, pp. 


Not only is the storm recognized, but the "picture" shown on the 
indicator is of sufficient detail to enable the forecaster to determine 
the area, speed, direction of movement, and intensity of any such 
storm within the range of the radar equipment. With a range of 
100-150 miles (160-240 km), warnings can be issued five or six 
hours prior to the arrival of the storm. When the storm is within 50 
miles, a good estimate can be made of the intensity and extent of 
the precipitation area. It is evident that the more accurate and ex- 
plicit forecasts (made possible by radar developments) have numer- 
ous applications to the protection of life and property on land, on 
sea, and in the air. In the absence of severe storms, radar echoes re- 
ceived from successive layers of clouds indicate the cloud levels 
and thicknesses. They thus supplement radiosonde observations and 
pilots' reports in the day-to-day charting and forecasting of the 

Another device for locating storms at a distance is "sf erics" equip- 
ment, abbreviated from atmospherics. This device is not radar; it is 
a static direction-finder, having receivers and amplifiers that pick 
up, and indicate the direction of origin of, the static electrical dis- 
charges that accompany thunderstorms and tornadoes. If the storm 
can be pinpointed on the radarscope, its location can be determined 
accurately by finding its direction from two or more base stations. 
Such equipment may soon prove extremely valuable in locating and 
tracking the dreaded tornado. 

In the case of high winds at sea, accurate observations of the 
height, period, and direction of the waves give valuable information 
on the location and intensity of the storms. Conversely, meteorolog- 
ical observations at sea are used to forecast for some time in advance 
the height and other characteristics of the sea waves and of the surf 
along the shore. 

Extended forecasts. In the preparation of extended forecasts for 
periods of from five to seven days in the future, three somewhat 
different methods of approach have been used by different groups 
of forecasters. These are discussed briefly in the following para- 

Method of five-day means. In forecasting by the use of five-day 
averages, reliance is placed mainly on the 700-millibar pressure con- 
tour chart and the circulation index to enable the forecaster to fore- 
see the pressure pattern that will exist during the next five to seven 


days. First, a chart is prepared on which the isobars show the aver- 
age pressure at sea-level for the five-day period ending the day be- 
fore the forecasts are made. Next, a similar chart is made on which 
the contour lines show the average height of the 700-millibar surface 
for the same period. These charts obscure the small and irregular 
fluctuations in pressure distribution and circulation index, but em- 
phasize their general trends and bring out the position and intensity 
of the centers of action. A comparison of these five-day means with 
the corresponding normals for the time of year is then made. 

From a study of these charts, the forecaster first makes a forecast 
of the index in the westerlies for the coming week and then prepares 
prognostic charts of the upper-air and surface mean-pressure distri- 
bution for the five-day period ending a week ahead. Finally, prog- 
nostic charts of surface pressure distribution are drawn for each day 
of the forecast period, and daily forecasts are made from these as 
from the usual daily synoptic charts. 

Method of extrapolation. An extrapolation method has been used 
in the preparation of daily forecasts for a six-day period beginning 
with the day following that on which the forecasts are made. In this 
method, a thorough study is made of existing conditions and tenden- 
cies as shown by the surface charts, and in particular by pressure 
contour charts at various surfaces in the upper air, extending into the 
stratosphere. Special attention is given to the two main types of cir- 
culation in the westerlies, corresponding to high and low values of 
the circulation index. Change charts are prepared to show past 
movements, both at the surface and in the upper air. Then the ex- 
pected positions of pressure centers, ridges, and troughs at the vari- 
ous levels at stated times in the future are extrapolated by extending 
previous movements and trends, step by step. Lastly, for each of the 
six days, the frontal systems and isobars are drawn, the air masses 
named, the areas of expected precipitation outlined, and the fore- 
casts stated. 

Method of weather types. A third method of approach used in 
the preparation of extended forecasts is called the weather-type 
method. By this method future weather conditions for a few days 
are determined by a study of past cases where similar conditions 
obtained (analogues). The method makes use of the fact that the 
weather in the zone of the westerlies is largely controlled by the 
irregular outbreaks of polar air along the polar front and by the 


varying position and character of the subtropical cells of high 

A large outbreak of polar air, for example, produces a typical 
meridional (north-south) circulation, which gradually becomes more 
and more zonal (westerly) as the polar air moves from its source 
and warms and merges with other air. This is the usual behavior of 
the circulation in the zone of the prevailing westerlies an occasional 
abrupt change from zonal to meridional flow, and then a gradual 
trend back to zonal in a period of a few weeks. A study of a long 
series of daily weather maps has shown, also, that three days is the 
most frequent interval between successive cyclone families or frontal 
passages, and that the principal troughs and ridges in the heart of 
the westerlies travel an average of 15 of longitude a day. 

In applying these principles, the zone of the westerlies is divided 
into geographical regions in each of which the major controlling 
factor is either in the subtropical cells or in the polar outbreaks. In 
each of these regions, various types of pressure distribution and 
frontal activity are set up. This is done by a study of past weather 
maps for a long series of years and by classifying them into types in 
which similar conditions obtain and similar weather is observed. 
In the North American region, about twelve principal types have 
been catalogued. 

Making use of these types and of such longer-term trends or tend- 
encies as may appear, the forecaster proceeds to determine the cur- 
rent type and then to predict the succeeding type on the assumption 
that past analogous situations will be repeated. Having done this, he 
prepares prognostic charts for each day of the forecast period, and 
from these prepares the daily forecasts. This method is not limited 
to a forecast period of a fixed number of days, but there is increasing 
error beyond five or six days. 

The preparation of the long series of weather maps of the North- 
ern Hemisphere, upon which this method of extended forecasting 
depends, occurred during World War II. It resulted in the accumu- 
lation of a file of modern daily surface weather maps of the Northern 
Hemisphere covering a period of forty years. These maps were care- 
fully classified and separated into types in terms of fundamental 
atmospheric processes and were then indexed and catalogued. An 
archive of classified weather maps, convenient for reference and re- 


search, was thus created. It was put to immediate use in the making 
of extended forecasts by the weather-type method, and has also been 
found useful in affording some indication of monthly and seasonal 
weather trends. It may well be the source of additional information 
and new concepts, as the effort to interpret atmospheric processes 

Long-range forecasts. Attention has been given for many years 
to the possibility of foreseeing weather conditions for a month or 
more in advance. Several different approaches to the problem have 
been developed. Attempts have been made to determine the general 
character of the weather a month, a season, or a year in advance by 
various statistical and physical approaches. These are discussed in 
Chapter 16, under the head of seasonal forecasting. In the present 
connection, we shall consider briefly the more strictly meteorological 
method of attacking the problem. This method in the main is an 
attempt to find types or analogues. 

Making use of methods similar to those used in the preparation of 
extended forecasts for a few days by the study of weather types, at- 
tempts have been made to extend the period farther into the future. 
In the classification of weather types, it is found that there are cer- 
tain trends, tendencies, or oscillations that show themselves as a rep- 
etition of a given type or a series of closely related types. For ex- 
ample, the cycle that follows an outbreak of polar air, from zonal to 
meridional flow and back again to zonal, is usually completed in 
from 25 to 40 days, averaging about 35 days. This has been called 
the polar sequence, or polar cycle. Such trends and oscillations are 
used in estimating future developments. 

However, in making such estimates, chief reliance is placed upon 
pressure and wind conditions in the upper troposphere and the 
lower stratosphere, as shown by upper-air charts. At these heights, 
conditions change slowly and, to a certain degree, predictably, It is 
assumed that these upper-air conditions steer or control surface 
conditions. The assumption is made also that future weather for a 
month or more will follow a course like that followed in similar situ- 
ations in the past, that the same weather sequences will occur. This 
is the essence of the method of analogues. 

By such means, attempts have been made, not only to predict the 
general character of the weather or the average weather for a month 


to a year in the future, but even to forecast the weather for each day 
for a month or more. This method of analogues has been used to 
some extent by our military forces, but it must still be regarded as 
in a preliminary, experimental stage, especially in the forecasting of 
daily weather for more than ten days in advance. No sufficient basis 
has been established for the assumption that one weather situation, 
which at best we can define only incompletely, always leads to the 
same subsequent weather, nor for the assumption that the course of 
events can be timed accurately. Continuous progress is being made, 
however, in measuring and interpreting the physical reactions of the 
atmosphere, and this advance keeps alive the hope for an ultimate 
understanding that will permit of accurate long-range prognosis. 

Numerical weather prediction. As early as 1910, L. F. Richard- 
son believed that weather forecasts could be made by solving the 
hydrodynamical equations of the atmosphere, thus eliminating the 
empirical aspects of synoptic meteorology. The equations are com- 
plicated, however, and about 14 years were required for Richardson 
to complete a 24-hour forecast. This obviously was not a practical 
technique. The development of high-speed electronic computers 
made the purely scientific approach to forecasting seem more prac- 
tical. Accordingly, a research project, originally sponsored by the 
Office of Naval Research but later also supported by the Air Force 
and Weather Bureau, developed numerical weather prediction to a 
degree of reliability comparable with existing forecast methods. 8 As 
a result of the success of this project, a numerical prediction unit was 
activated at Washington, D. C., in the summer of 1954. The fore- 
casting success of the electronic calculator depends on the accuracy 
of the data it receives and its ability to integrate, within a relatively 
short time, the interrelationships of atmospheric variables over the 
forecast period. 9 The numerical prediction technique is most prom- 
ising for increasing the accuracy of short-range weather forecasts of 
48 hours or less. If it is ever to be useful for long-range forecasting, 
the short-range forecast must increase to near 100 per cent accuracy. 

8 Philip D. Thompson, "An Introduction to Numerical Weather Prediction." Cam- 
bridge, Mass.: Mimeographed report of the Atmospheric Analysis Laboratory, Geo- 
physics Research Directorate, Air Force Cambridge Research Center, 1954. 

Louis Berkofsky, "A Numerical Prediction Experiment," Bulletin of the American 
Meteorological Society, Vol. 33 (Sept., 1952), pp. 271-273. 


Local Forecasts and Weather Lore 

"Men judge by the complexion of the sky 
The state and inclination of the day." 

Forecasts for a few hours in advance may be made without instru- 
ments or maps from the appearance of the sky, from the wind, and 
from the "feel" of the air. They are made by everyone who looks out 
in the morning and decides whether to carry a raincoat or not. "When 
clouds appear, wise men put on their cloaks." Farmers, sailors, and 
others who watch the weather closely may become quite adept in 
such forecasting. 

Wind-barometer indications. A barometer will help in making 
such local, short-period forecasts, but its indications are not simple, 
The words on the dial of an aneroid barometer mean little. It is not 
the actual reading of the barometer so much as the kind and rate 
of change of pressure that are of importance. The following rules 
for interpreting changes in wind and pressure are useful in the ab- 
sence of additional information and official forecasts: 

"When the wind sets in from points between south and southeast 
and the barometer falls steadily, a storm is approaching from the 
west or northwest, and its center will pass near or north of the ob- 
server within 12 to 24 hours with wind shifting to northwest by way 
of southwest and west. When the wind sets in from points between 
east and northeast and the barometer falls steadily, a storm is ap- 
proaching from the south or southwest, and its center will pass near 
or to the south or east of the observer within 12 to 24 hours with 
wind shifting to northwest by way of north. The rapidity of the 
storm's approach and its intensity will be indicated by the rate and 
the amount of the fall in the barometer/' 

Statistical indications. From a long series of observations, tables 
or graphs may be prepared showing the probability of rain or other 
weather occurrences under certain conditions of pressure, tempera- 
ture, or wind direction. Thus, it was found at Dubuque, Iowa, that 
during the summer months rain fell within 12 hours in 93 per cent 
of the cases when the following conditions were recorded at the 
morning observation, namely, pressure between 29.75 and 29.85 
inches (1,008 and 1,011 mb) and falling, temperature also falling 
and sky cloudy. The percentage was only 33 under conditions which 



were the same except that the pressure was rising and the sky clear. 
It was also found that the probability of rain within 24 hours was 
72 in 100 in all cases when the wind was from the east and the 
barometer falling, and only 44 in 100 when the wind was from the 
northwest. This and other results are shown in Fig. 118. 













Fig. 118. Percentage of Time Rain Occurred During June and July at Dubuque, 
Iowa, Within 12 and 24 Hours, as Related to the Wind Direction and the Barometric 
Tendency at the Morning Observation; Based on 3,390 Observations. A, pressure 
falling, rain within 24 hours; B, pressure falling, rain within 12 hours; C, pressure 
rising, rain within 24 hours; D, pressure rising, rain within 12 hours. 

Such results are local in their application, and studies of this char- 
acter cannot hope to take the place of synoptic charts in forecasting. 
They may, however, furnish supplementary information and sug- 
gestion, and, in the absence of a weather map, are of value in indi- 
cating probable local weather conditions. The best results for short 
periods in advance are obtained by combining the use of the weather 


map with a knowledge of local signs, such as are furnished by clouds, 
wind directions, and pressure changes. This is true for short periods 
only. For forecasts for 12 or more hours in advance, dependence 
must be placed on the weather map, for the weather can change 
greatly in that period, and it is often misleading to look out of the 
window and try to anticipate tomorrow's weather from today's. 

Single-station forecasting. It happens often in war and some- 
times in peace that there are isolated units on land or on ships at sea 
to whom a knowledge of coming weather is of much importance but 
to whom synoptic reports are not available. A meteorologist in such 
a situation can gain valuable information from a study of the clouds 
if he is thoroughly familiar with their various types and subtypes 
and the details of their structure. The amount of cloudiness and 
whether it is increasing or decreasing, and the shapes and character- 
istics of the clouds tell him much concerning the structure of the air 
masses in which the clouds lie. They help to identify the air masses 
and to indicate their degree of stability, and thus give an indication 
of the coming weather. A detailed knowledge of cloud characteristics 
is of importance at all times to the forecaster and the pilot, but espe- 
cially when only local data are at hand. 

If the isolated unit is equipped with pilot balloons and radio- 
sondes, or airplanes with recording instruments, additional informa- 
tion valuable in forecasting may be obtained. The pilot-balloon 
records of the direction and force of the wind at different levels give 
an indication of the positions of the centers of areas of high and low 
pressure, and the direction of the isobars and isotherms, at upper 
levels. From these indications, inferences may be drawn concerning 
the movement of pressure systems at the earth's surface. Radiosonde 
or airplane observations, especially if they extend to the tropo- 
pause, give information as to the value of the circulation index, the 
types of air masses at different levels, and the location of fronts. 
Clouds and local free-air soundings thus help the forecaster to get a 
fairly authentic picture of existing local weathet* conditions and prob- 
able future conditions, even in the absence of an extensive network 
of observations. 

Remembering that weather normally travels from west to east in 
middle latitudes, a single station may take frequent detailed observa- 
tions of the weather and plot each succeeding observation to the 
west of the preceding one. In the trade-wind belt of low latitudes, 


where weather moves from east to west, the plotting should he done 
in reverse order. After several entries have been made, the fore- 
caster has an approximate cross section of the atmosphere in one 
direction from his station. Thus he has a better concept of the pres- 
sure pattern, wind field, and general weather conditions than would 
otherwise be possible. This technique is especially useful on ships 
and small island stations where local orographic effects are neg- 

Weather proverbs. Some of the weather proverbs relating to the 
appearance of the sky, the direction of the winds, and the humidity 
of the air, are the result of long experience and have a certain valid- 
ity, similar to the statistical results mentioned above; but many of 
them lose their application when transplanted from the part of the 
world where they developed. None of them are "sure signs." For 
example, cirrus clouds often develop far in advance of a warm front, 
and when they thicken into cirrostratus, it is a good indication that 
the warm-front rain is approaching; but rain does not fall in the 
entire area over which the cirriform clouds appear. 

The conditions of plants and the condition and behavior of ani- 
mals are largely responses to past and present weather influences, 
but probably they furnish no indication of future weather. The belief 
in such omens is suggestive of the superstitious practices of the au- 
gurs of ancient Rome. It may be true that certain plants and animals 
are sensitive to changes in pressure, temperature, and humidity, and 
for that reason their actions just before a storm may give a few min- 
utes' warning of an approaching change. There is no "equinoctial 
storm" except in the sense that the equinoxes mark transition periods 
between winter and summer conditions, and storms often occur in 
the regular progress of the seasons during the latter part of March 
and of September. The long-range "forecasts" published in almanacs 
are generalized statements for large areas based on normals and 
normal variations and are confirmed only by chance. 

Weather control. The atmosphere is so vast and unconfined that 
efforts to control it over large areas seem almost hopeless. Man's 
accomplishments at rain-making are probably the nearest approach 
to weather control, yet they are so small in comparison with natural 
phenomena that their magnitude has defied accurate measurement. 

Everyone can recognize the desirability of being able to control 
the rainfall, prevent hail, or dissipate tornadoes and hurricanes, but 


these capacities remain as challenges for the future in the realm of 
atmospheric dynamics. Man has done a much better job in adjusting 
himself to the environment than in adjusting the environment to 


1. (a) Examine Fig. 87 and explain why more precipitation is likely 
to occur from cP air in summer than from cP air in winter. 

( b ) Is there any foundation for the old adage that it may get "too 
cold to snow"? 

2. Using the first of a series of daily weather maps: 

(a) Make a 24-hour forecast of temperature, wind, and pressure 
for a station near your home. 

(b) Note the position of all highs, lows, and fronts. 

3. Using the second of a series of daily weather maps: 

( a) Transpose the former positions of all highs, lows, and fronts to 
the new map. 

(b) Compute hourly velocities of the high and low centers. 

(c) Compute the speed of the strongest front on the map at three 
different places along lines normal to the front. 

(d) Make a 24-hour forecast for the movements of the pressure 
centers and the front. 

(e) Check your forecast for accuracy against the following map in 
the series. 

4. Make a list of the weather proverbs in use in your community. Try 
to trace each of them to their origin and to conclude whether they are 
based on fact or fantasy. 




In the use of the air as a medium through which to propel great 
ships of many tons' weight, some knowledge of the properties and 
behavior of the air is obviously necessary. The operation and safety 
of airplanes are not independent of atmospheric conditions, in spite 
of the fact that modern methods, instruments, and flying aids have 
greatly reduced flying hazards. 

Airways Service 

With the development of aviation has come a more intensive and 
extensive use of weather data. The Weather Bureau and the Civil 
Aeronautics Administration have developed, along the principal air- 
ways of the country, a service by means of which the officials at the 
airports and the pilots in flight are kept constantly informed of 
weather conditions along the routes traveled and are provided with 
specific forecasts of conditions to be expected during the next few 

Airways observations. Weather observations are made and re- 
corded hourly at hundreds of airways weather stations situated at 
airports in all parts of the country and at stations a considerable 
distance from the airways on both sides. At most stations, this is 
continued for 24 hours each day. In addition, special observations 
are made whenever marked changes in weather conditions occur. 
All these observations are transmitted over teletype circuits to re- 
gional centers and become available to all intermediate stations 
equipped with teletype service. They are known as sequence reports, 
or hourly sequences. From the regional centers, the existing weather 



conditions along the airways, as shown by these reports, are broad- 
cast at frequent intervals for the information of all, but especially 
for pilots in flight. Thus, both officials and pilots at the airports and 
pilots in the air are kept continuously informed of current weather 
conditions, The broadcasts are available to all planes equipped with 
radio-receiving facilities. A network of weather-reporting stations is 
indispensable in the regular operation of airways. 
The elements of the sequence reports are: 

( 1 ) Station identification. Identifying code letters are assigned 
each reporting station by the Civil Aeronautics Administration, 

(2) Type of report. This group is omitted except for a special 
report which indicates important weather changes, 

(3) Ceiling. Ceiling and cloud heights are expressed in hun- 
dreds of feet. The manner in which the ceiling was determined is 
indicated by A, aircraft; B, balloon; , estimated; M, measured; and 
W, indefinite. 

(4) Sky condition. Sky condition is reported as clear, scattered 
clouds, broken, or overcast. The sky may also be reported as ob- 
scured by dust, fog, haze, or smoke. 

(5) Visibility. Visibility is reported in statute miles and frac- 

(6) Weather. Letter symbols are assigned to indicate prevailing 
weather conditions and obstructions to vision. 

(7) Sea-level pressure. Three coded digits indicate tens, units, 
and tenths of millibars. 

(8) Temperature. Temperature is reported in degrees Fahren- 
heit. A minus sign ( ) is prefixed when the temperature is below 

(9) Dew point. Dew point is reported in the same manner as 
temperature and is especially valuable when fog or icing conditions 
are anticipated. 

( 10) Wind direction and speed. The wind is reported by arrows 
indicating direction, followed by its speed in miles per hour. 

(11) Altimeter setting. Three coded digits expressing units 
tenths, and hundredths of inches allows the pilot to adjust his pres- 
sure altimeter so that it will read the proper elevation above mean 
sea level. 

(12) Remarks. Used to report in plain language, symbols, or 


contractions significant weather phenomena not included elsewhere 
in the message. 
A sample sequence report is decoded below: 

PIT SI M3V QD 7 1%V L-FK 146/72/68 -> 11/989/CIG 2V4 VSBY 
1V2 Pittsburgh, Pa., special report no. 1, measured ceiling 300 feet 
and variable, broken clouds with overcast at 700 feet, visibility l l / 2 miles 
and variable, light drizzle with fog and smoke, pressure is 1014.6 milli- 
bars, temperature is 72F., dew point is 68F., west wind at 11 miles per 
hour, altimeter setting is 29.89 inches, ceiling variable from 200 to 400 
feet, and visibility variable from one to two miles. 

Airways forecasts. Weather service to aviation was made a pri- 
mary responsibility of the Weather Bureau by federal legislation in 
1938. The service was reorganized in 1952 as the "Flight Advisory 
Weather Service," whose function includes all weather forecast and 
warning services to air navigation, provided by 26 FAWS forecast 
centers. Forecasts contain information on clouds, cloud heights, visi- 
bilities, fog, haze, smoke, wind conditions, icing, turbulence, the lo- 
cation and intensity of local storms and frontal systems, and other 
weather elements of particular interest to aviation. 

Aviation forecasts are of four types: (1) regional forecasts, which 
cover large geographical areas; (2) terminal forecasts, which are 
individual forecasts of probable weather conditions at air terminals; 
(3) trip forecasts issued upon request to cover an individual flight; 
and (4) advisory service to Civil Aeronautics Authority traffic con- 
trollers. In addition, special international forecasts are prepared to 
serve planes flying beyond the boundaries of the United States. 

The principal air transport companies employ their own meteor- 
ologists, who co-operate with the government officials and supple- 
ment the official forecasts by special advice to their own pilots at the 
beginning of each flight. Clearance must be given to the pilot by the 
company meteorologist and is given only when he knows the weather 
conditions and thinks them safe, not only at the destination but also 
at intermediate and surrounding airports. In the making of forecasts, 
use is also made of observations received from airplanes in flight and 
of the regular upper-air observations. The United States Air Force 
and Navy make similar airways forecasts for the guidance of their 
own personnel in addition to making use of the regular forecasts. 

Besides the hourly sequence reports and the six-hourly forecasts 
and summaries, the pilot has access before take-off to direct informa- 


tion concerning conditions above the surface. Pilot-balloon and 
rawin observations give him the direction and speed of the wind at 
various levels in the atmosphere. Radiosonde observations provide 
data on the temperature, pressure, and relative humidity of the air 
at short intervals from the surface up to the stratosphere. Graphs, or 
sounding curves, showing the vertical structure of the air, are con- 
structed on adiabatic charts from these radiosonde reports. They 
supply the data also for upper-air charts showing temperature and 
pressure distribution and wind conditions at fixed levels above the 
surface and for charts showing weather conditions at constant pres- 
sure levels. 

Flying Weather 

The existence of such a complete organization for the collection 
and distribution of weather information is evidence of the impor- 
tance of weather in aircraft operation. It follows that the pilot 
should have enough knowledge of meteorology to interpret prop- 
erly the reports and forecasts and to understand the significance of 
any changes that he may encounter while in the air. He needs to 
know the characteristics of air masses, of warm and cold fronts, and 
of cyclones and anticyclones. He should know under what condi- 
tions to expect unusual turbulence and when ice is likely to form 
on exposed surfaces. He should recognize quickly the various cloud 
types and their methods of formation, and know when to attempt 
to fly over them, under them, or through them. A knowledge of 
the average frequency of different wind directions and speeds and 
their changes with altitude is important to the pilot. The wind may 
be used to advantage in flying, since it is frequently from different 
directions at different levels. Some of the more important weather 
conditions that are sources of danger to aviators are discussed 
briefly in the following paragraphs. 

Thunderstorms. Thunderstorms present a major hazard in flying 
because of the violent vertical movements of the airup-and-down 
movements in close proximity to each other. Upward velocities 
sometimes reach 100 miles per hour. When a plane passes from such 
an updraft to a strong downdraft, there is danger of structural 
damage and damage to flight instruments. Under such circum- 
stances, it is extremely difficult for the pilot to maintain control of 


his plane. Turbulence may cause the instruments to oscillate so 
much as to make accurate readings impossible. For these reasons, 
thunderstorms are to be avoided if at all possible. The turbulence 
danger is worst in the lower forward portion of the cumulonimbus 
cloud, and it is particularly important to avoid the squall cloud. 
High wind velocities extend to the top portion. The top is often 
from 25,000 to 35,000 feet (8-10 km) in height and sometimes 
penetrates to 70,000 or 80,000 feet (22-25 km) above sea level. Many 
planes are not designed to fly over the tops of cumulonimbus, nor 
are they equipped with the necessary oxygen supply to fly at these 

Air-mass thunderstorms are generally scattered and occur more 
frequently by day. They can usually be avoided. Frontal and pre- 
frontal thunderstorms are harder to avoid because they frequently 
occur in long, continuous squall lines. They are usually more severe, 
also, especially along and ahead of a strong cold front. Cold-front 
thunderstorms are low-level storms, often attended by a large area 
of low ceiling and low visibility. They are also attended by rapid 
changes in pressure, necessitating frequent resettings of the alti- 

When the air moving against a mountain slope is conditionally or 
convectively unstable, mountain thunderstorms may be severe and 
continuous and may extend to great heights, presenting dangerous 
or impossible flying conditions. A thunderstorm presents not only 
the hazards of excessive turbulence, but also those of hail, lightning, 
icing, and low visibility. Hail damage is not serious under ordinary 
circumstances, but occasionally large hailstones do damage aircraft 
while in flight, A lightning stroke is very likely to put the electrically 
operated equipment out of operation but otherwise is not dangerous 
to airplanes of metal construction; it may seriously damage fabric, 
plastic, and plywood construction, however. Near a thunderstorm, 
static may interfere with radio communication. 

Icing of aircraft. The accumulation of ice on the exposed parts 
of airplanes in flight through clouds, fog, or precipitation is one of 
the most serious weather hazards of aviation. If observations show 
high humidities and steep lapse rates, icing conditions are probable 
at heights where temperatures are below freezing. Icing ordinarily 
occurs only in the presence of supercooled water droplets in the air. 
These freeze rapidly when they strike a solid object. Supercooled 



drops are found even at temperatures of 20F., but are most fre- 
quently encountered at temperatures between 15F. and 32F. 
Temperatures of 26F. to 32F. are particularly dangerous, for the 
air can contain more moisture at these temperatures than at lower 
temperatures. Any clouds with temperatures between 0F. and 32F. 
should be avoided as probable sources of icing. Icing occasionally 
occurs when a cold airplane enters an area of falling rain that has 
a temperature somewhat about freezing. The drops may then freeze 
upon striking the cold surfaces of the airplane. 

Two principal types of ice deposit occur, namely, rime and clear 
ice (or glaze}. When very small supercooled drops strike the air- 
plane, they do not spread, but freeze as small pellets, forming a 
rough, granular surface, milky in appearance. This is called rime 
(Fig. 119). It is usually not deposited rapidly enough, nor in suf- 
ficient weight, to be a serious danger. It adheres loosely and is fre- 
quently dislodged by vibration. 


Fig. 119. Rime and Glaze on Wing of Aircraft. Drawing by E. L. Peterman. 

Larger drops break upon contact with the aircraft and spread 
backward to form a surface of clear ice, which adheres strongly to 
solid surfaces and therefore is much more dangerous than rime. 
There is special danger when flying through, rain or dense clouds 
with the air temperature not far below 32F. In such circumstances, 
the ice may build up rapidly, even as much as two or three inches 
in as many minutes. This icing rapidly adds weight, reduces the 
lift, and increases the vibration and the stalling speed. The rate of 
deposit depends upon the size and the number of the drops en- 
countered, and therefore, on the speed of the plane and the type, 



density, and temperature of the cloud. Polar Pacific air masses show 
a higher frequency of icing than do other air masses in the United 

In stratiform clouds, the air is usually stable, the drops are small, 
and icing takes the form of rime. In most cases, it is possible to avoid 
such clouds by flying over them. In cumuliform clouds, the warm, 
unstable, rising air causes the drops to grow to large size and car- 
ries them into temperatures below freezing. In such clouds, icing 
is often rapid and dangerous. Fortunately, these clouds are often 
scattered and the pilot can go around them. At times, when an air- 
plane passes from cold air into a cloud that is not so cold but is 
below freezing, there is a rapid formation of frost over the entire 
outside portion of the plane. This occurs most frequently in clouds 
with ascending currents. It is not so dangerous as the usual types of 
icing, but it reduces cruising speed and increases stalling speed. 

The first impulse of a pilot, after severe icing has reduced air 
speed or caused loss of altitude, is to climb above the icing condi- 
tion. This may be impossible or impractical and creates an additional 
hazard when the angle of attack is changed very much. The ice 
then builds up in irregular fashion, thus destroying the air foil, an 
effect which further increases the drag, reduces the lift, and adds 
total weight to the plane. 1 Ice may also bother the pilot by accumu- 
lating on the propeller, windshield, or air-speed indicator. Ice, snow, 
or frost is especially dangerous on a plane at take-off, and should be 

Fig. 120. Icing in Frontal Conditions. A, cold front; B, warm front. 

* Robert W. Mudge, Meteorology for Pilots. New York: McGraw-Hill Book Co., 
1945, pp. 115-144. 


removed, especially from the wings, before the plane attempts to 
leave the ground. 

Icing is frequent in the frontal zones of both warm and cold 
fronts. It is frequently rapid and severe in the cold wedge underly- 
ing the warm air in advance of a surface warm front, when rain is 
falling from the warm air (Fig. 120). The drops become super- 
cooled in the cold air or freeze upon contact with the cold aircraft. 
Snow falling through supercooled drops is especially dangerous, for 
the snow adheres and adds to the accumulated weight. It also 
roughens the surface and increases drag. Rime or clear ice, or both, 
may form also in the cloud area above the inversion in a warm front. 
The safest plan for the pilot is to fly over the clouds of a warm front 
when the temperatures indicate that icing is probable. 

At a cold front, convection is active and icing will probably be 
rapid to severe along the squall line. Flying through a cold front 
should be avoided if possible; but, if it is necessary, the course 
should be perpendicular to the line of the front. Severe icing may 
occur in flying through the clouds formed in the rising air on the 
windward side of mountains, even though mountain thunderstorms 
have not developed. Conditions are especially hazardous if cumulus 
clouds are forming. 

Small planes are also subject to carburetor icing, which may be 
more dangerous than any other type because it is not visible and 
can occur in summer as well as in winter. It is believed that car- 
buretor icing is responsible for more engine failures of light planes 
than any other single cause. 

The carburetor mixes the fuel with air and meters the mixture 
to the engine. To insure proper mixing and vaporization of the fuel, 
a venturi (constriction) is built into the air intake at the point of 
the fuel jet (Fig. 121). This increases the speed of the air past the 
fuel jet as it rushes into the partially vacuumized area created by 
the intake stroke of the pistons. Rapid expansion of the air, in re- 
sponse to the reduced pressure, cools it adiabatically. Additional 
cooling results from vaporization of the fuel. The temperature of 
the air may be reduced 40 to 50F. in this manner, causing mois- 
ture to freeze about the throttle. Carburetor icing, under ideal con- 
ditions, may accumulate rapidly and cause engine failure with very 
little warning. It may be avoided by preheating the air before it 
reaches the venturi. 






Fig. 121. Carburetor Icing. Air pressure is reduced on the engine side of the venturi, 
causing rapid cooling by expansion. Cooling is further accelerated by vaporization of 
the liquid fuel. 

Fog, visibility, ceiling. A third major weather hazard in aviation 
is fog, in its relation to visibility and ceiling. Fog and low stratus 
clouds are among the most frequent causes of flying accidents. Fly- 
ing conditions are poor if the horizontal visibility is one mile or less, 
or if the ceiling is 500 feet or less. Conditions are only fair when 
visibility is between one and three miles and when ceiling is be- 
tween 500 and 1,500 feet. Fog often produces these conditions. Radi- 
ation ground fogs may be dense near the ground, but are shallow. 
In most cases, these fogs have their greatest vertical extent and their 
lowest visibility about sunrise. 

Ground fogs are local in character, often forming as a result of 
air drainage in lowlands and in river valleys when they do not form 
over higher ground in the same vicinity. One airport may be closed 
by ground fog while a near-by airport is entirely clear. Hence, when 
flying at night or in the early morning, the pilot must be on the look- 
out for fog at his terminal station, even though there is none at 
stations en route. The rate of fall of temperature during the night 
gives a means of estimating when the temperature and dew point 
will come together and the fog begin. Advection fogs are deeper, 
more extensive, and more persistent than are ground fogs, and, 
therefore, are not so easily avoided. They may persist throughout 
the day. Visibility may be reduced dangerously low, also, by haze, 
dust, smoke, and blowing snow, and by falling rain or snow. 

Low ceilings caused by low stratus clouds are of frequent occur- 
rence. The pilot loses sight of the ground and must resort to in- 
strument flying. If he is flying low, he runs the risk of striking the 
ground. The elevation of the base of a stratus layer, and hourly 
reports of ceiling height, are of fundamental importance to the avia- 
tor. The other forms of low clouds ( stratocumulus and nimbostra- 



tus) frequently form ceilings close to the earth, and ceilings under 
clouds with vertical development (cumulus and cumulonimbus) are 
often less than 1,500 feet. 

In emergencies, planes may be able to land safely when the visibil- 
ity or ceiling is near zero. The pilot must be skilled in instrument 
flying and have the necessary instruments at his disposal. He may 
choose the ILS (instrument landing system) approach if his plane 
is equipped with an automatic pilot capable of accurately bring- 
ing the plane in on a radio beam. A more widely used landing 
method at commercial and military fields is called GCA (ground- 
controlled approach). Essentially, its success depends on the skill 
of a GCA controller to watch the plane's progress on a radarscope 
and direct the pilot by radio to a safe landing approach ( Fig. 122 ) . 

2 3 45 6 7 

Fig. 122. GCA Radarscope. Shows vertical and horizontal position of the plane as it 
approaches the landing field. 

A GCA landing requires only that the plane be equipped with radio 
facilities, and even small private planes have been successfully 
landed by this method. 2 

2 John R. Hoyt, "Flying the GCA," Flying Vol. 54 (May, 1954), pp. 30-31, 63. 


Still another technical landing aid was announced by the U.S.A.F. 
in 1953. It is known as VOLSCAN and is reported to be able to pick 
up planes several miles from the traffic circle and unerringly direct 
them to a safe landing approach at the rate of 60 to 120 planes per 
hour. 8 These navigational and landing aids have rendered fog, 
smoke, low ceiling, and other obstructions to vision of much less 
consequence to aviation. 

Turbulence. The wind is never steady; it is always moving in 
irregular gusts of varying speed and direction, as has been noted in 
previous chapters. The degree of turbulence varies (1) with the 
speed of the wind and its inherent irregularities of flow, (2) with 
the roughness of the surface over which it is moving, and (3) with 
the stability of the air. Turbulence is extreme in and around thun- 
derstorms and is great enough to be dangerous in active cumuli- 
form clouds and in high winds blowing against large, abrupt ob- 
structions. Turbulence is ordinarily light in stratiform clouds and 
where steady rain is falling, and it is usually light at heights of 5,000 
feet or more above the ground. 

The pilot should be completely familiar with the different types 
of clouds, their manner of formation, and their range of elevation 
and thickness. They are visible evidence to him of what is happen- 
ing in the atmosphere and of its degree of turbulence. In military 
operations, clouds are often used for concealment. In these cases, 
it is especially necessary that the pilot know the clouds and be able 
to estimate the comparative dangers of flying through the cloud 
and of enemy action. 

Turbulence on the windward side of mountains in air moving 
upslope is ordinarily not dangerous unless thunderstorms or cumu- 
lus clouds have developed. Strong winds moving down the leeward 
sides have irregular eddies and may cause trouble, especially in 
landing. The downdraft on the lee side also causes a loss of alti- 
tude, and the pilot must approach the mountain well above it in 
order to avoid crashing into it (Fig. 123). In strong winds, the 
eddies around buildings situated on or near an airport may result 
in accidents on landing or taking off. 

Stability reduces turbulence by checking vertical movement. 

8 "Computer Times Final Approaches," Aviation Age, Vol. 21 (January, 1954), 
pp. 44-49. 




Fig. 123. Eddy Turbulence in Mountains. A plane flying upwind in mountainous 
country requires more altitude for safe flight than when flying in the opposite 

When there is a temperature inversion in the lower air, the marked 
stability contributes to quiet, nonturbulent conditions, but it also 
contributes to the development of fog at the ground and a ceiling 
of stratus clouds at the inversion level. Local, daytime instability 
over a ground surface that is being heated by the absorption of 
the sun's rays often adds to the turbulence caused by the speed of 
the wind. This occurs even though the instability is not sufficient to 
produce cumulus clouds. 

Differences in the heating of the earth's surface over different 
areas result in such turbulent movements and produce what is 
known as bumpiness. Bumps occur near changes in the character 
of the surface, as along the borders of forests, and where cultivated 
and grass-covered areas meet. Even such narrow strips as roads 
and creeks are often marked by bumps, with ascending currents 
over roads and descending currents over creeks. Bumpiness due to 
such irregularities of surface heating is ordinarily greatest within 
the first 2,500 feet (765 m) above the ground. Although unpleas- 
ant, it is usually harmless to airplanes in flight. A somewhat differ- 
ent type of bumpiness is encountered in the waves occurring at the 
surfaces of different air layers, and these layers may be at any 

A knowledge of the characteristics of warm and cold fronts and 
of the different air masses of which they are composed is of much 
importance to the pilot. A large number of accidents occur at or 


near cold fronts because of the violent action and rapid changes 
attending such fronts. Warm fronts are usually more stable and less 
turbulent, but they present hazards of fog, low visibility, and icing. 
Visibility is generally better in polar than in tropical air masses be- 
cause of the greater stability and lower moisture content of the polar 
masses. Most precipitation falls from tropical maritime air masses, 

Pressure-pattern flying. An example of the combination of me- 
teorological knowledge with instrumental improvement and flying 
skill is found in the development of what is called pressure-pattern 
flying. This procedure was recently developed to enable aircraft to 
maintain regular, round-trip flight schedules across the stormy North 
Atlantic Ocean during the winter months. Hitherto, this route, al- 
though the shortest, had been avoided in winter, especially for west- 
bound flights against the prevailing westerlies. Owing to the stormi- 
ness of the North Atlantic in winter and to the absence of surface 
stations over wide expanses, serious weather hazards not previously 
reported were often encountered in flight during World War II. 
The plan evolved to overcome these difficulties depends upon the 
use of the radio altimeter, itself a wartime product, in combination 
with the usual, pressure-actuated altimeter. 

The pilot may fly a constant altitude by the use of his radar alti- 
meter and make a record at intervals of the altitude indicated by 
the pressure altimeter. Likewise, he may fly at constant pressure by 
the use of the pressure altimeter and note the difference in eleva- 
tion indicated by the two instruments. In either case, the indicated 
differences in height may be converted to corresponding pressure 
values. By plotting these values for readings made at frequent in- 
tervals, the pilot gets a map of the pressure distribution at the level 
of his flight. With this up-to-date pressure pattern before him, he 
is able to compute the pressure gradient and to determine his posi- 
tion relative to the dominant pressure center. He is able to avoid 
hazards and take advantage of favorable conditions. By this means 
regular, all-year, round-trip flight schedules across the North Atlan- 
tic were established. 4 Pressure-pattern flying is now being practiced 
by commercial airlines over continental areas also. 

4 Howard E. Hall, "Use of 'Pressure-Pattern Flying* Over the North Atlantic," 
Bulletin, American Meteorological Society, Vol. 26, 1945, pp. 160-163. 



1. Assume that you are the official observer for an air field, and make 
an observation for a sequence report at your locality. 

2. Construct a visibility scale for your locality by following these in- 

(a) At the center of a sheet of cardboard, construct a circle one- 
half inch in diameter. 

( b ) Draw nine more concentric circles about the first, each having 
a radius one-half inch greater than the preceding one. 

(c) Mark the points of the compass about the outer circle. 

(d) Using a scale of one-half inch equals one mile, draw in sym- 
bols for several of the most distinct objects at varying distances 
and in various directions from your position. 

3. Can there be carburetor icing when the dew point is above 32 F.? 
Use your knowledge of adiabatic processes to explain your answer. 

4. Visit a commercial air field and ask to see aircraft deicing equip- 
ment. * 

5. While at the air field, you should visit the weather station. Try to 
read the hourly sequence reports. 




Certain electrical and optical occurrences in the atmosphere are 
of general interest in connection with meteorology, although for 
the most part they have little relation to what is ordinarily meant 
by weather. The most frequent electrical display in the air and the 
one most directly related to weather is lightning, which has been 
discussed in connection with thunderstorms. This chapter presents 
a summary discussion of other important electrical and optical phe- 
nomena of the atmosphere. 1 

Electrical Phenomena 

Electrical field of the earth. In the discharge of a lightning flash 
from a cloud to the earth, there is evidence of a great difference of 
electrical charge between the cloud and the earth, but it is not only 
in a thunderstorm that such a difference exists. There is a continual 
difference of potential between the earth and the atmosphere. Usu- 
ally the earth is negative as compared with the air, but reversals 
frequently occur in thunderstorms. The potential gradient differs 
from place to place, at different times of the year, and at different 
times of the day. The difference is greatest in fall and early winter 
and least in summer. 

Conductivity of the air. This electrical charge on the earth con- 
tinues even though there is a gradual loss of charge through the air. 

1 For more detailed descriptions and physical explanations, the reader is referred to 
W. J. Humphreys, Physics of the Air, 3rd ed., New York: McGraw-Hill Book Com- 
pany, Inc., 1940; and B. F. J. Schonland, Atmospheric Electricity, New York: John 
Wiley and Sons, Inc., 1953. 



Although the air is a poor conductor of electricity, it is not a per- 
fect insulator, because of the presence of free ions. The number of 
ions in the atmosphere, and hence its conductivity, change with 
changing weather conditions. 

As a consequence of the conductivity of the air, there is a current 
between the earth and the atmosphere sufficient to neutralize the 
earth's charge in a short time. It is thought that a preponderance 
of negative electricity in lightning flashes to the earth and a pre- 
ponderance of positive electricity in the brush discharges of pointed 
conductors connected with the earth help to maintain the earth's 
negative charge. Whether these are quantitatively sufficient, or 
whether there are other compensating currents, is not known. There 
are known to be frequent irregular currents in both directions. The 
best-known example of a brush discharge is called St. Elmo's fire, 
and consists of short streamers of light appearing at the ends of 
pointed objects, especially on mountains. -* 

Auroras. The northern lights, Aurora Polaris, are luminous phe- 
nomena associated with the earth's magnetic* field in the region 
around the north magnetic pole. The conspicuous but irregular na- 
ture of the display of lights has attracted the attention of scientists 
for many centuries. Auroral displays are closely related to magnetic 
disturbances on the earth and to sunspot activity on the sun. An 
increase in the number of auroras has been correlated with the 11- 
year interval of maximum sunspots. A much shorter 27-day cycle of 
auroral maximia coincides with the movement of sunspots around 
the solar axis. It is believed that auroras result from the interac- 
tion of the earth's magnetic field with showers of electrical cor- 
puscles thrown off by the sun. The light is caused by the excitation 
of gas molecules, primarily oxygen and nitrogen, in much the same 
way that rarefied neon is made luminous in familiar lighting tech- 

The aurora of the Northern Hemisphere is called aurora borealis 
and is visible most frequently in a zone which crosses Alaska, north- 
ern Canada, southern Greenland, Iceland, and northern Norway. 
The frequency decreases rapidly with decrease of latitude. Auroras 
are less frequently observed from the northern United States and 
only rarely from southern Europe. 

There is a definite nocturnal distribution of auroral frequency. 
Recent studies have shown a concentration of auroral occurrences 


about one hour before magnetic midnight (midnight at the north 
magnetic pole). Some auroral activity also occurs by day but is not 
visible to the naked eye. 

Fig. 124. Aurora Band Just After Sunset, Looking Westward at Bossekof, Norway, 
March 3, 1910. (Courtesy, U. S. Weather Bureau.) 

Auroral displays take the form of arcs, bands, rays, draperies, 
coronas, crowns, or luminous fog-like patches (Fig. 124). A single 
aurora may remain quiet, move forward and backward in wave-like 
motion, or shoot spectacular streamers diverging across the sky 
from a single point. Those showing ray-like structure are especially 
active and colorful. Most arcs, bands, and patches are primarily 
green in color, but the rays often show red, yellow and violet. The 
height of auroras above the earth varies generally from 60 to 200 
miles (100 to 300 km). A few auroras have been observed in the 
polar region to be only a few hundred meters above the ground, 
and some rays have been observed from Norway to reach a height 
of 465 miles (750km). 

Auroral displays also occur in the Southern Hemisphere, where 
they are known as aurora australis. Observation and study of the 
aurora australis has been limited by the absence of land in the lati- 
tudes most favorable for viewing the phenomena. 


Optical Phenomena 

Refraction. In passing from one medium to another of different 
density, waves of radiant energy are refracted, or bent from their 
original straight-line course, except when they are traveling per- 
pendicularly to the surface separating the two media (Fig. 125). 
The bending is explained by the fact that the waves travel at dif- 
ferent speeds in media of different densities. The function of lenses 
in microscopes, telescopes, and eyeglasses is to bend light out of its 
course. A straight stick partly immersed in water appears to be bent, 
because the light by which we see it is refracted at the water's 

White light is composed of waves of different lengths, and the 
longer waves are refracted less than the shorter ones. When white 
light is passed through a glass prism, it is separated into colored 
bands. In the spectrum that is thus produced, there is a regular 
gradation of colors, but for convenience seven colors are often dis- 
tinguished and called primary colors. These are red, orange, yellow, 
green, blue, indigo, and violet, in the order of their wave lengths, 
red being the longest. Light is also bent on passing through air of 
changing density. In this case, the refraction is too diffuse to pro- 
duce a separation into colors; it results in a gradual change in the 
direction of the wave. 

, V 



Fig. 125. Refraction of Light on Passing from One Medium to Another of Different 
Density. The wave front travels more rapidly in the thinner medium and is bent out 
of a straight-line course. From A, an object at B appears to be at B'. 

Atmospheric refraction. One effect of refraction in the air is to 
bend the rays of light coming from the sun, moon, or stars, when 


they are near the horizon, into a curved path which renders these 
objects visible when they are, whether rising or setting, in reality 
about a half-degree below the horizon. It is the increasing density 
of the lower air that causes the rays to be bent downward and gives 
to distant objects a greater apparent than actual elevation. A ray of 

Fig. 126. Atmospheric Refraction Resulting from the Air's Increasing Density 


light from A, Fig. 126, follows the curved path AO to the observer 
at O and appears to come from A'. The effect of this refraction with 
respect to the sun is to lengthen the time between sunrise and sun- 
set by a few minutes in lower and middle latitudes and by more 
than 24 hours in polar regions during a part of the year. Another 
effect is to flatten the disk of the sun or moon when it is on the 

The twinkling of the stars is also a refraction phenomenon, re- 
sulting from the movement of masses of air of different densities 
across the line of sight. Sometimes when the air is free from visible 
dust and condensed moisture, it has a haziness which renders dis- 
tant objects indistinct. This occurs especially on hot days, when 
convection is active. It is called optical haze and, like the twinkling 
of the stars, is due to the unequal refraction of light passing through 
air of varying density. 

Mirage. In addition to the usual results of atmospheric refrac- 
tion just described, special optical effects known as mirages are oc- 
casionally seen. These occur when there are strong temperature con- 
trasts in adjacent air layers. The most common forms of mirages are 
those in which there are deceptive appearances of water surfaces 
and those in which there are images of distant objects. The com- 
mon inferior mirage, producing the illusion of a water surface, is 
often seen in flat desert regions on quiet, sunny days and also, on 



a small scale, on paved roads (Fig. 127). It is due to a thin, heated 
surface layer of air, perhaps 3 or 4 feet in thickness, with consider- 
ably cooler and denser air above it. This is a condition favorable 
for convection, but in which convection has not yet begun, because 
of lack of turbulence. The apparent water surface is the image of 
the sky. The eye of the observer must be somewhat above the heated 
layer. In this type of mirage, a distant object and its inverted image 
some distance below it are sometimes seen. Because the image is 
below the real object, it is called an inferior mirage. 

Fig. 127. Mirage in Death Valley, California. Courtesy, U. S. Weather Bureau. 

Over water and over cold land surfaces, a cold layer of air at the 
surface and warmer air above it sometimes produce a superior 
mirage in which a distant object and two images above it are seen. 
For instance, one may see a ship in the distance, and apparently 
another ship above it floating upside down, and above this still an- 
other ship, upright in this case. These phenomena are due to gradual 
bending of the light. We speak of layers of warm and cold air, but 
it should be understood that the transition from one layer to an- 
other is not abrupt. There is mixing and a gradual change in the 
refractive power of the air, and the effects seen in mirages are due 
to this continuous variation. Complete stratification, with mirror- 
like bounding surfaces, does not occur in nature. 

Rainbow. When a ray of sunlight enters a drop of water, a part 


of it does not pass directly through but is reflected from the inner 
surface and emerges on the side from which it entered, being re- 
fracted both on entering and on leaving the water (see Fig. 128). 
In this way are produced the concentric colored arcs of the primary 
rainbow, having a radius of about 42, with the red on the outside. 
A portion of the light, however, is twice reflected before emerging, 
producing the secondary bow, with a radius of about 50, in which 
the red is on the inside. It is usually not possible to distinguish all 
the primary colors because of overlapping of the spectra from dif- 

Fig. 128. Path of Light in Producing a Primary Rainbow, (0), and a Secondary 

Rainbow, (b). 

ferent drops of water. The larger the drops, the narrower and 
brighter is the bow produced. 

The observer of a rainbow is always between the sun 1 and the 
falling rain and in the line connecting the sun and the center of 
the circle of which the rainbow is an arc. Each observer sees a dif- 
ferent rainbow, that is, a bow made by different drops of water. If 
the sun is on the horizon, it is possible to see a complete semicircle; 
if the sun is more than 42 above the horizon, no primary rainbow 
is possible. Lunar rainbows are sometimes seen. 

Halo. When light from the sun or moon passes through thin 
upper clouds composed of ice crystals, various circles or arcs of light 
may become visible. These are solar or lunar halos, respectively, and 
are produced by refraction. The most common halo is a ring, or 
portion of a ring, of 22 radius, around the sun or moon (Fig. 129). 
(The angular diameter of the sun and of the moon is about %.) A 
halo of 46 radius is sometimes seen, and occasionally there are 
other circles and arcs making complex figures, in which reflection 


also plays a part. The various figures are due to differing shapes and 
positions of the falling ice crystals through which the light passes, 
Halos are often nearly white, but a well-developed halo of 22 is red 
on the inside, shading off to yellow. 


Fig. 129. Some of the More Frequently Observed Halo Phenomena. HH, horizon; 
S, sun; Z, zenith; aa, halo of 22; bb, halo of 46; cc, parhelia of 22; dd, portion of 
parhelic circle; ee, upper tangent arc of halo of 22; /, sun pillar. (After W. J. Hum- 

Parhelion. A parhelion is a bright spot, or mock sun, occurring 
on the parhelic circle, usually where it intersects another halo. The 
parhelia most frequently observed are at or near the intersection of 
the parhelic circle with the halo of 22 (Fig. 130). They form two 
colored spots, often called sun dogs, one on each side of the sun, and 
are red nearest the sun. The corresponding image in connection 
with a lunar halo is called a paraselene. 

Diffraction and scattering. Light rays are diverted from a 
straight course, made to curve around a corner, as it were, when 
the obstacles encountered are of a size comparable with the wave 
lengths of light. This phenomenon is called diffraction of light and 
results in the formation of alternate light ai}d dark or colored bands. 
Diffraction is the cause of the corona, a bluish-white aureole some- 
times visible immediately around the sun or moon, when either of 
these is veiled by a thin cloud composed of small water droplets. 
The outer rings of a corona often show some rainbow-like coloring. 
It is of value to an aviator to distinguish between a halo, which 
shows the presence of ice crystals, and a corona, which shows the 
presence of water droplets. 



Fig. 130. Halo of 22 Around the Sun and Parhelic Circle Through the Sun. Courtesy, 

U. S. Weather Bureau. 

The molecules of the gases of the air, and also the fine dust par- 
ticles that have a diameter less than the wave length of the light, 
cause scattering, a diffusion or dispersion of light somewhat similar 
to diffraction. Scattering of light decreases the intensity of direct 
sunshine but greatly increases the brightness of the sky away from 
the sun and the amount of light received by objects in the shade. 
Scattering and diffraction are responsible for the blue color of the 
sky and for the changing, often brilliant, colors of twilight. Over- 
head in a clear sky, the short, blue waves penetrate the air; near 
the horizon, the light must pass through long distances of the dusty 
lower air, and only the longer waves toward the red end of the 
spectrum succeed in getting through. That these colors are due to 
the influence of the lower air is shown by the fact that space ap- 
pears nearly black when viewed from the stratosphere. 

Twilight. The existence of the phenomenon of twilight is due 
in part to diffraction and scattering and in part to reflection of sun- 
light from the upper atmosphere. Some light from the upper air 
reaches us more than an hour before the sun is above the horizon in 


the morning (more than two hours at some seasons in northern por- 
tions of the United States), and some light continues for a like 
period in the evening after the sun has set. This is the period of 
astronomical twilight, when the sun is not more than 18 below the 
horizon. The time that it takes the sun to move 18 toward or away 
from the horizon depends upon the angle which the sun's apparent 
path makes with the horizon, and hence upon the latitude and the 
time of year. In general, it is too dark for outdoor work when the 
sun is more than 6 below the horizon. Hence, the interval between 
the time when the sun is on the horizon and when it is 6 below 
is called civil twilight. The duration of this interval also varies with 
the latitude and the time of year. It averages about 22 minutes at 
the equator, about 29 minutes at latitude 40, and about 37 minutes 
at latitude 50. 


1. Place a coin on the bottom of a pan and then walk slowly backward 
until the coin ceases to be visible over the rim of the pan. Without mov- 
ing, have someone pour water into the pan until you can again see the 
coin. Explain this phenomenon. 

2. Would it make any difference where you would need to stand in 
Problem 1 if alcohol were used? Would it take more or less alcohol in 
the pan to produce the same effect as water? 

3. Stretch a string tightly around a globe to represent the boundary 
between daylight and darkness on March 21st. Let the string pass through 
your home town as if the sun were setting. Now cut a long strip of paper 
that is as wide as 18 of latitude on your globe. Place the strip of paper 
along the "darkness" side of the string to represent twilight. Can you 
figure how long twilight will last in your town? 

4. Use a nautical almanac for sun position and adjust your string for 
today's condition. Compute the length of twilight and check your results 
against actual observation. 




The sciences of meteorology and climatology inevitably overlap. 
No sharp line can be drawn between them. To discuss the weather 
adequately, one must consider the frequent, usual, or average 
weather conditions in different parts of the world, and these are 
what comprise climate. To understand climate, one must know some- 
thing of the reasons for the various kinds of weather experienced, 
for climate is the total effect of all the daily weather. In the study of 
meteorology it is desirable, therefore, to consider also some of the 
main features of the climates of the earth and their relation to, and 
influence upon, one another and upon man and his manner of life. 

Climatic Elements 

The elements of weather and climate are so numerous and com- 
bine in such endless variety that the complete description of a cli- 
mate is extremely difficult. Accurate description involves the exten- 
sive use of climatic tables and extensive statistical analysis to reveal 
the characteristics of the data. To be complete, the tables should 
include all the elements that affect man and his activities. To acquire 
from such tables a correct idea of what a given climate is like re- 
quires not only a careful comparison of the data with similar data 
for climates with which one is familiar, but also the exercise of some 

Climatic data. Temperature and rainfall are the two most im- 
portant climatic elements, but the simple tabulation of their aver- 
age annual values is not sufficient. Two places with the same mean 
temperature and rainfall may have very different climates because 



of differences in the distribution of these two elements within the 
year. For example, San Francisco and St. Louis each has a mean an- 
nual temperature of 56, but San Francisco's January average is 
49.9, and St. Louis's is 31.1; whereas in July the averages are 58.5 
and 78.8, respectively. Some of the more important data used in 
the description and classification of climates are: 

Mean monthly and annual temperatures, and mean annual ranges 
of temperature. 

Mean daily maximum and minimum temperatures, and mean 
daily ranges. 

Highest and lowest temperatures of record. 

Average number of days with maxima above 90F.; above 100F. 
Average number of days with minima of 32F. or lower; of 0F. or 

Mean monthly and annual relative humidity. 

Mean monthly and annual precipitation. , 

Greatest and least monthly and yearly amounts of precipitation. 

Greatest precipitation in 24 hours. 

Excessive amounts of rainfall for short periods. 

Average snowfall by months. 

Average number of days with rain, snow, hail, fog, thunder. 

Mean cloudiness in tenths of sky covered. 

Mean monthly frequency and duration of low ceilings. 

Mean percentage of sunshine by months. 

Mean wind velocity by months. 

Prevailing wind direction by months. 

Mean frequency of winds from the different directions. 

Mean frequency of gales. 

Mean frequencies of the different types of air masses. 

Average and extreme dates of first and last killing frosts. 

Analysis of climatic data. A broad, generalized statement of the 
character of a climate in terms of mean, totals, and extremes is read- 
ily obtained by examination of the data recorded under the head- 
ings listed in the preceding paragraph. This has long been the 
standard way of describing climates. It has served well to picture 
the main features of a particular climate and to distinguish the major 
types of climate. Military operations, both at the surface and in the 
air, however, require more specific and detailed climatic information. 

The need for such information calls for a new type of statistical 


treatment that would make possible an accurate calculation of 
weather risks. Not only the frequencies of individual elements, but 
also the combined frequencies and interrelations of two or more 
elements, or of the same element at two or more points, are required. 
Such problems as the following are illustrative: frequency and dura- 
tion of fog in the English Channel for each wind direction each 
month; frequency and duration of low ceilings at a given airport 
with reference to the time of day, or the direction and force of the 
wind, or with reference to simultaneous conditions at other airports. 

The solution of such problems by examining the records of indi- 
vidual stations and making the necessary tabulations and calcula- 
tions by ordinary statistical methods is impracticable because of the 
time and labor required. This fact, together with the urgent need 
for such information, led to the introduction of electrical sorting 
and tabulating machines using punched cards. These machines take 
over the entire job of sorting, recording, and adding the numerical 
data. They quickly furnish a new type of climatological statistics, 
not previously available. After the data have once been entered by 
punching holes in standardized cards in accordance with a stand- 
ardized system, the arrangement and computation for any desired 
element or combination of elements can be done at the rate of 10,000 
cards per hour. The cards become a permanent depository for fu- 
ture studies of many kinds. 

The punched-card method had previously been used to a limited 
extent for special climatological studies both in the United States 
and in Europe, but its wide application to all sorts of climatic prob- 
lems is now beginning and marks a fundamental development in 
climatological methods. Punched cards will make much valuable 
specific information quickly available to a large number of human 
activities. To have definite information on the probability of favor- 
able or unfavorable climatic conditions at a given time of year and 
with reference to specific crops or farm activities will be of obvious 
advantage to the farmer. A knowledge of the frequency of certain 
weather conditions and combinations of weather conditions is also 
of direct value in transportation by land, sea, or air, in the design 
and construction of residences, warehouses, and other buildings, in 
the construction of airports, in the management of forests, in flood 
control and irrigation practice, and in many other fields. All these 


activities could then be planned with more confidence on the basis 
of calculated weather risks. 

Punched-card analysis has been used also to ascertain the relations 
between the types of weather occurring at the same time in differ- 
ent parts of a large area, such as a state. If a certain pressure dis- 
tribution pattern shown on weather charts is attended by rain in 
western Oregon, for example, what are likely to be the weather con- 
ditions in eastern Oregon at the same time? What is the probability 
of clear weather, of overcast skies, Or of rain? 

To answer such questions, a long series of daily climatological 
records covering the entire area under consideration may be ana- 
lyzed to determine the percentage frequencies of weather condi- 
tions occurring simultaneously in different parts of the area. The 
results obtained constitute a synoptic climatology of the region. The 
method was developed and used primarily in the planning of large- 
scale military operations but has valuable applications in other fields. 
It aids the forecaster by contributing to an understanding of weather 
processes, and it furnishes important additional detail in the descrip- 
tion of a climate. 

Air-mass climatology. Another new and developing treatment of 
climatic data is known as air-mass climatology. By this method, 
climates are examined and described with reference to the preva- 
lence of the various types of air masses and the influence of each in 
giving a place its particular climate. This requires, first, a record at 
each observing station of the air mass or masses present during each 
24 hours, and of the time of passage of fronts. From an air-mass 
calendar obtained in this way, the relative frequencies and durations 
of the various air masses may be computed by months and seasons. 
For example, a knowledge of the relative frequency of polar and 
tropical air masses and of continental and maritime air masses gives 
new and important information concerning the climate of a region. 
Not much has yet been done in the field of air-mass climatology 
because of the lack of the fundamental data. Daily records of air 
masses were not begun at United States Weather Bureau stations 
until July, 1945. Hence, the records are still too short to serve as 
bases for normal values. 

Solar and physical climates. If the earth were a uniform land 
surface without an atmosphere, the temperature of the surface at 


any given place would be governed directly by the amount of inso- 
lation received there. The annual amount of insolation is greatest 
at the equator and least at the poles, and, under the conditions 
assumed, we should have a regular decrease of temperature from 
equator to poles, The actual air temperatures over the earth, as it 
is, follow this plan of distribution in main outline but not in detail. 
Insofar as the climate of a place depends directly on the amount of 
solar radiation received, it is called solar climate. 

The division of the earth into the five classic zones bounded by 
the Tropics of Capricorn and Cancer and by the polar circles is of 
ancient origin and is purely on a solar-climate basis. These are zones 
of possible sunshine rather than of actual climate. Within the 
Tropics, the sun is vertically overhead at noon twice each year. 
Within the polar zones, the sun is below the horizon for at least 24 
consecutive hours in winter and above for at least 24 hours in sum- 
mer. In the intermediate zones, the sun is never in the zenith and 
never below the horizon for 24 hours. The latitudinal zones, not- 
withstanding their old names of torrid, temperate, and frigid, merely 
mark differences in the elevation of the sun. 

The actual or physical climate does not follow the parallels of 
latitude. It is modified by geographic conditions, chiefly ( 1 ) by the 
irregular distribution of land and water, (2) by winds and ocean 
currents, and ( 3 ) by differences in elevation. These modifying influ- 
ences act in various ways to produce climatic differences: land and 
water absorb and radiate heat differently; cloudiness and humidity 
are influenced by distance from large bodies of water; movements 
of air and water convey large amounts of heat across latitudinal 
lines. Nevertheless, the distribution of insolation is the primary fac- 
tor determining temperature. Solar climate is the groundwork upon 
which modifications are imposed by other factors. 

Distribution of Temperature 

Except in polar regions, the normal distribution of temperature 
over the earth is now fairly well determined. Most inhabited land 
areas have temperature records of moderate length. Although rec- 
ords over the oceans are less extensive, nearly all vessels at sea make 
regular observations, and, because ocean temperatures are less vari- 


able than land temperatures, these serve quite well to determine 
the general temperature distribution. The distribution of tempera- 
ture is indicated on a map by lines drawn through points of equal 
tempertaure, isothermal lines. 

For the daily weather maps and for maps of small areas, the actual 
temperatures are usually represented, but on maps of extensive 
areas, where there are great differences of level, mean temperatures 
are first reduced to sea level by using a lapse rate about equal to the 
average lapse rate in the free air. This procedure is necessary if the 
lines are to show the effects of latitude and of continental land 
masses on the distribution of temperature. If the actual tempera- 
tures obtained at different altitudes were used, these more general 
influences would be obscured; besides, it is not possible to indicate 
on a small-scale map all the temperature differences found in moun- 
tainous regions. Where the isotherms shown on the maps presented 
in this chapter pass over elevated regions* they do not represent the 
actual temperatures to be found there but have been thus corrected 
for altitude. 

Normal yearly temperatures. The first and most obvious fact 
noted on examining a world chart of mean annual temperatures is 
the decrease of temperature from equatorial regions toward polar 
regions (Fig. 131). This decrease is evidently due to the different 
amounts of insolation received; solar climate is dominant in deter- 
mining the general course of the isotherms. They do not follow the 
parallels of latitude closely, however, but bend irregularly north- 
ward and southward. In equatorial and lower middle latitudes, they 
bend poleward over the continents, indicating that the continents 
in these latitudes have an average temperature warmer than that of 
the oceans, or than the average of the latitude around the globe. In 
Siberia and northern Canada, the isotherms bend southward, show- 
ing that large continental areas in high latitudes are colder than the 
adjacent oceans. 

The isotherms turn far northward in the north Atlantic, disclosing 
the influence of the warm water that moves from our Florida coast 
northward and then northeastward across the Atlantic. A similar 
though less marked northward trend occurs in the Pacific from Japan 
to Alaska, along the course of the Kuroshio and its extension, the 
North Pacific Current (West Wind Drift). In the Southern Hemi- 



sphere, cold ocean currents flow northeastward toward the west 
coasts of South America and Africa, bending the isotherms equator- 

There are thus two major influences producing the irregularities 
of the annual isotherms: (1) the differing responses of land and 
water to the influence of insolation, and (2) the transportation of 
warm and cold water by ocean currents. Note how these influences 
result in a crowding of the isotherms in Alaska and southeastward 
to New England, and in eastern Asia. The warmest area, as ex- 
pressed by the annual means, is in central Africa, where the tem- 
perature averages more than 85. The isotherm of 80 extends around 
the world except for small areas in the eastern Atlantic and eastern 
Pacific Oceans, and includes large portions of Central and South 
America, Africa, Arabia, India, and Indo-China, all of the East In- 
dies and the Philippine Islands, and a part of northern Australia. 
The coldest regions of the world cannot be shown so definitely be- 
cause records are very incomplete for polar regions, especially for 
Antarctica. There is a short record from central Greenland which 
gives a mean annual temperature of 5F., and a record of one year 
at Framheim on the coast of the Ross Sea in Antarctica gave an an- 
nual mean of 14.4F. 

January and July normal temperatures. An examination of the 
January and July temperature charts (Figs. 132 and 133) discloses 
the migration of the isotherms with the seasons. The January iso- 
therm of 90F. includes only small areas in south Africa and in Aus- 
tralia. In July the average temperature is 90 or over in a part of 
southwestern United States and large areas in north Africa and 
southwestern Asia. A small area in the Colorado Valley has a July 
mean of 95, and a portion of the Sahara desert, a mean of 100. In 
contrast to these high temperatures, the temperature of interior 
Greenland remains below freezing throughout the year. In January 
the lowest mean temperatures are in Siberia, 50F., and in Green- 
land, 40 F. The lowest mean shown on the map for July is 20F. 
along the border of the Antarctic continent, but for August, 1911, a 
mean of 49F. was obtained at Framheim. 

Note also the migration of the isotherm of 70 in North America; 
in January it crosses Mexico, and in July it has moved northward to 
Canada. The change of temperature is less over the oceans than over 





the lands, and, hence, less in general in the Southern Hemisphere 
than in the Northern. The migration as a rule is less than that of the 
sun's rays, which is 47 of latitude. As a result of continental and 
oceanic influences, modifying the effects of insolation, the January 
temperature off the coast of Norway is 40 higher, while in the 
interior of North America and Asia it is 30 lower, than the lati- 
tudinal average, In July these anomalies, or departures from the 
average temperature of the latitude, are generally less than 10F., 
except in the interior of the United States, Asia, and north Africa, 
where they are from 10 to 20. 

Annual and daily ranges of temperatures. Mean annual ranges of 
temperature are an expression of the average difference between the 
mean temperature of the warmest and of the coldest month. They 
are shown in Fig. 134. It will be seen that mean annual ranges are 
much greater over continental interiors than over large ocean areas 
in the same latitude, except in equatorial regions. They naturally 
increase with increase of latitude because of the greater difference 
between winter and summer insolation as the distance from the 
equator becomes greater. In the tropical oceans and across equa- 
torial Africa and South America, the annual range is less than 5F. 
It increases to 30 F. near the tropics in Africa, South America, and 
Australia, to 80 in the interior of Canada, and to 120 in a small 
area in Siberia. This progressive change shows clearly the effects 
of distance from equator and distance from unfrozen oceans. 

In the chart of January mean daily ranges of temperature in the 
United States ( Fig. 135 ) , the influences of humidity and of eleva- 
tion are clearly evident. In portions of the elevated, arid southwest, 
the average difference between day and night temperatures in Janu- 
ary is 33F., while in the vicinity of Puget Sound, the marine influ- 
ence results in a daily range of only 9F. Note the influence of the 
Great Lakes in reducing the range. Two physical principles are in- 
volved in this influence: first, the water is slow to change its tem- 
perature; second, the increased humidity of the air screens out some 
insolation by day and absorbs earth radiation by night. Note that the 
extent of the marine influence on the Pacific coast is greater than on 
the Atlantic. 

Some extremes of temperature. On page 315 are some records 
of extreme temperatures, obtained under standard conditions of ex- 
posure and expressed in the Fahrenheit scale: 




Lowest temperature, -90F., Verkhoyansk, Siberia, Feb. 7, 1892. 

Lowest average monthly mean temperature. 60F., Oimekon, Siberia. 

Lowest temperature in the United States, 70F., Rogers Pass, Mon- 
tana, Jan. 20, 1954. 

Lowest mean temperature for one month in the United States, 13F., 
St. Vincent, Minnesota. 

Lowest temperature in Canada, 81F., Snag, Yukon Territory, Feb. 3, 

Lowest temperature in Alaska, 76F., Tanana, in 1886. (A minimum 
thermometer, left for 19 years near the top of Mt. McKinley, not in an in- 
strument shelter, when recovered indicated a minimum of approximately 

Lowest mean for one month in Yukon Territory, Canada, 51.3F., 
Dawson, December, 1917. 

Highest temperature of record, 136F., Azizia, Tripoli, Sept. 13, 1922. 

Highest temperature in the United States, 134F., Death Valley, Cali- 
fornia, July 10, 1913. 

Highest mean temperature for one day in the United States, 108.6F., 
Death Valley, California. * 

Highest normal annual temperature in the United States, 81.1F., Fort 
Stockton, Texas. 

General Distribution of Precipitation 

Precipitation occurs at irregular intervals and is greatly variable in 
amount, so that many years of record are required to obtain smooth 
daily, or even monthly, normals. In fact, it may be questioned 
whether the word normal in this connection has much significance. 
Records on land are sufficiently numerous and of sufficient length, 
however, to justify the use of mean values as tentative normals, with 
the understanding that several hundred years of record might alter 
them materially. There is little exact knowledge of the average 
amounts of rainfall over the oceans. On a chart representing the 
distribution of rainfall, lines of equal rainfall are called isohyets. 
Isohyets are drawn to indicate the actual precipitation; there is no 
reduction to sea level as in the case of isobars and some isotherms. 

Normal annual precipitation. From a map of the average annual 
rainfall of the world we may deduce the following general state- 
ments, which the reader should verify by reference to Fig. 136. 

1. Precipitation is greatest in equatorial regions and decreases 
irregularly toward the poles. The decreasing amount of moisture 
in the air as the temperature declines from equatorial to polar re- 
gions naturally results in a smaller total precipitation. Also, the gen- 



eral tendency of air to expand and rise in warm areas and to settle 
in cold areas leads to greater precipitation in the former as com- 
pared with the latter regions. 

2. Rainfall decreases toward the interior of large continental 
masses, because the chief source of supply of the moisture of the air 
is the oceans. Much of the moisture is often precipitated on the 
near-by land areas, and little is left for the distant interiors. Note 
the large dry areas in the central portions of Asia and North Amer- 
ica. ( Other factors are involved besides the inland position of these 
areas.) Over large land areas there is also an important secondary 
source of atmospheric humidity in the evaporation from lakes, rivers, 
soil, and vegetation. 

3. Rainfall shows a relation to the general wind systems of the 
world (discussed in more detail in the next section) and to the di- 
rection of the wind, especially whether onshore or offshore. 

4. Ocean currents influence the distribution of rainfall. Warm 
currents increase precipitation on the neighboring coasts, for there 
is much water vapor over warm water, and this vapor is cooled 
when it moves inland, as on the eastern coasts of North and South 
America. Cold currents diminish precipitation, for the air moving 
over them is cool and stable and of moderate humidity, as on the 
western coasts of South America and south Africa. 

5. Mountain systems influence precipitation by giving rise to 
ascending and descending air currents. Most mountain systems have 
a wet and a dry side, the wet side being toward the ocean or toward 
the prevailing winds. Outside of the tropics, the wettest parts of the 
world are mountain slopes facing prevailing winds from the oceans. 

Seasonal variation of precipitation. The various types of seasonal 
distribution of rainfall have great economic significance. There are 
large areas in equatorial regions where the rainfall is heavy through- 
out the year, and other areas within the tropics with alternate wet 
and dry seasons. In the middle latitudes, the west coasts of con- 
tinents have a winter maximum of rainfall and dry summers. The 
precipitation is cyclonic in origin and is often increased by oro- 
graphic factors. In the interiors there is a marked summer maximum, 
largely of thunderstorm type. On eastern coasts there is a fairly 
even distribution through the year, partly cyclonic and partly con- 
vectional, but usually with a summer maximum. 

The relation of rainfall to the growing season is of particular im- 



portance. For example, in a large portion of the Mississippi and 
Missouri Valleys, where the total precipitation is light to moderate, 
the heaviest rainfall occurs in the first half of the growing season, 
May, June, and July, when it is of the greatest value in the produc- 
tion of crops. A few types of monthly distribution of rainfall are 
shown in Fig. 137. 

Fig. 137. Types of Rainfall Distribution by Months. 

Annual number of rainy days. The relation of the total rainfall 
to the number of days with rain is a climatic factor of some impor- 
tance, indicative of the type of rainfall and of the general impres- 
sion of dampness or dryness given by the climate. In some places 
the rain falls in moderate or heavy showers of short duration, and the 
skies are clear for long intervals. These conditions are characteris- 
tic of the interiors of continents and of such regions as our Gulf and 
south Atlantic coasts. ( See Fig. 138. ) 

In other places there are many days of light rain or drizzle, giving 
a large number of rainy days but only light or moderate rainfall. 
In this country the coasts of Oregon and Washington have climates 


of this character, and in Europe the British Isles, the Netherlands, 
Belgium, and western France have similar conditions. In both con- 
tinents, these are marine climates in the prevailing westerlies. The 
region around the Great Lakes has a similar climate in this respect. 
Seattle has 151 rainy days and a normal annual rainfall of 34.03 
inches, giving an average of 0.23 inch per day of rain; Oklahoma 
City has nearly as much rain, 31.15 inches, but it falls on 82 days at 
the rate of 0.38 inch for each day of rain. Marquette, Michigan, has 
only 32.47 inches per year, but there are 165 days on which a meas- 
urable amount falls, each day on the average receiving only 0.20 
inch; at Pensacola, Florida, a much heavier rainfall, 57.85 inches, 
occurs on fewer days, 114, and the amount per day is 0.51 inch, or 
2% times the amount at Marquette. 

Areas of heavy and light rain. The average annual precipitation 
is above 100 inches in small areas in Central America, Panama, 
western Colombia, and southern Chile; in the East Indies, the Him- 
alayas, and along the north coast of the Gulf of Guinea. These are 
all warm regions, but profitable use cannot be made of the land, 
because the rainfall is too heavy, and the growth of native vegeta- 
tion is too luxuriant. Average amounts between 80 and 100 inches 
occur at places on the west coast of North America from Alaska to 
Oregon, in tropical South America, many tropical islands, and large 
areas of the tropical oceans. 

At the other extreme, there are areas of less than 10 inches in 
southwestern United States, the Sahara arid Arabian deserts, much 
of interior Asia from the Caspian Sea to China, the trade-wind belts 
of the eastern Atlantic, and in north polar regions north of latitude 
70. In the Southern Hemisphere there are regions of less than 10 
inches in South America, in southwest Africa, and in much of in- 
terior Australia. 

Amounts of rain between 20 and 100 inches are favorable for 
agricultural use of the land. Areas receiving between 10 and 20 
inches of rain a year are semiarid. They are suitable for grazing and 
dry farming, but not for intensive agriculture except under irriga- 
tion. Where the rainfall is below 10 inches, desert conditions exist, 
and water for irrigation must be brought from wetter regions. Pro- 
duction depends upon the yearly distribution of the rain, and upon 
other factors, notably temperature. The amounts, 10 to 20 inches, 



just given, are used as approximate dividing values. The following 
table shows approximate percentages of land areas of the earth with 
rainfall between given values: 



Annual Rainfall 
in Inches 

of Land Area 


Less than 10 












More than 80 


Very wet 

Some extremes of rainfall. Mt. Waialeale, Kauai, Hawaii (alti- 
tude 5,075 feet ) holds the record of being the wettest place on earth, 
having received an average annual rainfall of 471 inches over a pe- 
riod of 20 years. Not far behind is Cherrapunji, India (altitude, 
4,309 feet), with an average annual record of 428 inches. Prevail- 
ing winds, together with orographic lifting, account for the unusual 
rainfall at both stations, While Mt. Waialeak receives copious rains 
during all months, Cherrapunji receives nearly all of its precipita- 
tion during the .summer monsoon. During 1860-1861, within a single 
12-month period, Cherrapunji received 1,041.78 inches. The Wynoo- 
chee Oxbow, Washington, receives an average annual rainfall of 
144 inches. 

Precipitation extremes during a single month: Manoyuram, India, 264 
inches; Helen Mine, California, 71 inches. 

Precipitation extremes during a single 24-hour period: 45.99 inches at 
Baguio, Luzon, Philippines, July 14-15, 1911; 38.20 inches at Thrall, 
Texas, September 9, 1921. 

Some of the most unusual rainfall amounts ever recorded in the United 
States are shown in Table IV. 


Station (inches) 

Opid's Camp, Calif 65 

Pensacola, Fla 1.34 

Taylor, Tex 2.00 

Galveston, Tex 3.95 

Guinea, Va 9.25 

Holt, Mo 12.00 

Rockport, W. Va 19.00 

D'Hanis, Tex 21.50 

Smethport, Pa. 30.70 

Thrall, Tex 34.60 

Time required 

Date of 






April 5, 1926 



May 2, 1937 



April 29, 1905 



June 4, 1871 



Aug. 24, 1906 



July 22, 1947 



July 18, 1889 



May 31, 1935 



July 18, 1942 



Sept. 9, 1921 

1 David M. Ludlum, ed., Amateur Weatherman's Almanac. Philadelphia: Franklin 
Institute, 1952, p. 9. 


In Romania there is a record of a fall of 8.07 inches in 20 minutes, 
and at Puerto Bello, Panama, a fall of 2.48 inches in 5 minutes. In 
the Sierra Nevada in east-central California, Tamarack (altitude, 
8,000 ft. ) has an average seasonal snowfall of 451 inches. The great- 
est fall recorded there in a single season is 884 inches, the greatest 
monthly fall is 314 inches, and the greatest depth of snow on the 
ground at any time is 454 inches. 

At the other extreme, the average annual rainfall in inches is 1.33 
at Helwan, Egypt; 1.45 at Greenland Ranch, Death Valley, Califor- 
nia; 1.84 at Aden, Arabia; 4.16 at Arequipa, Peru; 5.39 at La Serena, 
Chile. There are considerable areas in southeastern California, west- 
ern and southern Nevada, and extreme western Arizona where the 
rainfall is less than 5 inches a year. 

Zonation of Climates 

A broad general description of the climates of the earth may be 
made, following a more or less latitudinal division into zones. Each 
zone includes many climatic variations, but some general charac- 
teristics applying to large areas may be mentioned. Although there 
are many contributing elements to the over-all climate of any area, 
such as solar radiation, temperature, rainfall, humidity, wind veloc- 
ity, and evaporation, temperature and rainfall are the most impor- 
tant for general consideration. 

Tropical zone. A simple climate is characteristic of the zone 
within the tropics. Its central portion is the equatorial low-pressure 
belt. This equatorial zone has a large annual rainfall and frequent 
and heavy thunderstorms in all months of the year. It is the doldrum 
region of variable winds and calms, a region of dense tropical forests 
of rapid growth. Bordering this wet belt on the north and on the 
south, there are regions which receive rain during the summer of 
that hemisphere, as the doldrums migrate toward the regions, but 
little or no rain during a short period of the opposite season, when 
the doldrums are farthest away. The vegetation of this climatic 
regime constitutes the true jungle, with many trees but also a dense 
undergrowth of tangled vines and other tropical plants. In the West- 
ern Hemisphere, these jungles extend intermittently northward into 
southern Mexico and southward into central Brazil. Toward the pole- 
ward sides of these belts, where the rainfall becomes light, there 


are open grasslands, or savannas, bordering the forests. The savan- 
nas include the Sudan of Africa, the Llanos of Venezuela, the Cam- 
pos of Brazil, and the Downs of Australia. In the equatorial and 
savanna zones, seasonal temperature changes are slight. There are 
practically no seasons, except where there is a wet and a dry season. 
Temperatures average high throughout the year, and the climate is 
oppressive and enervating, especially when the humidity is high, but 
maximum temperatures are usually not so high as in continental 
interiors in so-called temperate zones. In large areas of the tropics 
temperatures never reach 100. High humidity, dense vegetative 
cover, and days that are shorter than summer days in higher lati- 
tudes, are factors in keeping the maximum temperatures moderate. 

In the central portions of the trade-wind belts, on the poleward 
sides of the savannas, the winds blow with great regularity at a mod- 
erate speed, storms are very rare, and temperatures are uniformly 
mild. There are no frosts; the climate is tropic&l, Although the trade 
winds move for long distances over the oceans, they move from 
colder to warmer regions and are therefore rather dry except when 
there is orographic uplift. On the windward sides of highlands 
athwart the constant trades, the rainfall is heavy and frequent at 
all seasons. In other situations, the skies are bright, sunshine is abun- 
dant, and rainfall is light. Though not stimulating, this trade wind 
climate is comfortable and healthful in contrast with the mugginess 
of equatorial climates. 

Subtropical zones. In the poleward portions of the trade winds 
and the equatorward portions of the subtropical high-pressure belts, 
there are transition zones in each hemisphere with subtropical types 
of climate, not entirely free of frost. On the west coasts of the con- 
tinents in these latitudes is the Mediterranean climate. These areas 
have moderate temperatures throughout the year, with moderate 
rainfall in winter under the influence of the westerlies, and with dry, 
sunny summers under the influence of the subtropical belts of high 
pressure. The Mediterranean climate is of greatest extent in the 
countries bordering on the sea from which it is named, but there 
are small areas of this type of climate in California, and in South 
Africa, southern Australia, and central Chile. 

The east coasts of continents in these transition areas have a 
humid subtropical climate, more nearly continental in character 
than is the Mediterranean type, with greater annual temperature 


ranges, and with no dry season. In parts of southeastern Asia and 
the Netherlands Indies, a humid, monsoon climate prevails, having 
a short moderately dry season but a heavy annual rainfall from 
onshore winds. 

In the main, the subtropical zones are deficient in rainfall, with 
large arid and semiarid areas. We have noted that the trade winds 
are naturally dry. The high-pressure belts are dry because of sub- 
siding air, but they are subject to more variable winds and to occa- 
sional invasion by storms from the prevailing westerlies. The natu- 
ral dryness of these belts becomes evident when we note that the 
belts include most of the great desert areas in each of the five con- 

Intermediate zones. The middle zones of each hemisphere are 
regions of the prevailing westerlies, much interrupted and confused 
by local conditions and by traveling disturbances resulting from the 
meeting of polar and tropical air masses. Wide temperature ranges 
and marked changeableness of weather are striking characteristics. 
There is much variability of rainfall, which is generally heavier on 
the coasts and lighter in the interiors of the continents. In Russia 
and Siberia, in middle latitudes, there are large, unwooded, grassy, 
semiarid plains called steppes. These and similar regions in Hungary 
and the Great Plains of the United States have what is called a 
steppe climate. Such a climate occurs only in large continental in- 
teriors. There are, however, large inland areas within the interme- 
diate zones that have adequate rainfall, as do, also, the coastal 

Polar zones. About 8 per cent of the earth's surface is included 
in the polar zones, in which only a minimum of plant and animal 
life exists. There is almost continuous sunlight for a short time in 
summer, with some warm days, but the season is so short that the 
ground remains permanently frozen except in a thin surface layer. 
Precipitation is light. 

Zones of temperature. The earth may be divided into climatic 
zones by using isotherms instead of parallels of latitude. On this 
basis Supan has made the following divisions: 

1. Hot belt: the area inclosed by the mean annual isotherm of 
68F. This belt is irregular, mostly in the Northern Hemisphere, 
and somewhat larger than the torrid zone. The poleward boundaries 


represent approximately the limit of the trade winds and of the 
growth of palms. 

2. Cold caps: the area around the poles inclosed by the isotherm 
of 50F. for the warmest month. This isotherm represents the limit 
of the growth of cereals and of forest trees. In cold regions it is the 
temperature of the summer rather than of the year that determines 
habitability and vegetative growth. Hence the temperature of the 
wannest month is used instead of the average annual temperature. 

3. Temperate belts: the area between the hot belts and the cold 
caps. The northern temperate belt extends north of the Arctic circle 
in Alaska and in Eurasia, but the cold cap extends south of the 
circle in the Bering Sea and on the Labrador coast. In the Southern 
Hemisphere, the temperate belt reaches no farther south than lati- 
tude 55. 

For a more detailed study of climate, each of these belts may be 
divided into numerous subdivisions. Many sifch subdivisions have 
been made, both on the basis of temperature and amount and dis- 
tribution of rainfall, and on the basis of plant growth. 

Climate as Related to the Physical Features of the 
Earth's Surface 

The differences of climate so far discussed have been closely re- 
lated to distance from the equator, but there are climatic variations 
wholly independent of latitude. The elevation of an area and its po- 
sition relative to continents, oceans, and mountain systems give the 
area certain climatic characteristics in whatever part of the world it 
may be. Other factors, indirectly related to latitude but producing 
independent effects on climate, are the influence of prevailing winds 
and of ocean currents and the prevalence of cyclonic storms. 

Continental climates. In the interiors of continents the climate 
is usually rather dry and clear; that is, rainfall is light to moderate, 
relative humidity is low, and sunshine is abundant. Within the 
tropics, temperature contrasts are small over large land areas as well 
as over the oceans. In middle latitudes, continental climates are 
marked by severe winters and hot summers; in polar regions, the 
winters are long and severe and the summers short and cool. Steppe 
climates are dry continental, and a desert is an extreme type of dry 
continental climate. 


Marine climates. The climate of the oceans and of lands that are 
largely influenced by ocean conditions, islands, for instance, is char- 
acterized by small daily and yearly ranges of temperature, with 
nights and winters relatively warm and days and summers cool, 
That is to say, these climates are equable and moderate in their 
changes. Because water warms slowly, the springs are late and cool; 
because water cools slowly, the autumns are late and warm. In the 
interior of the United States, July is the hottest month, on the aver- 
age; at San Francisco, where the climate is largely marine, Septem- 
ber is the warmest month. Except in the trade wind belts, marine 
climates usually have greater humidity and cloudiness than con- 
tinental climates. 

Coastal or littoral climates. The climate along the coasts of con- 
tinents is intermediate between the marine and the continental 
types. The prevailing winds and mountain barriers largely determine 
the distance inland to which oceanic influences penetrate. In the 
zones of the prevailing westerly winds, west coasts of continents 
have belts of distinctly coastal climate; but on the east coasts, con- 
tinental climates extend practically to the shore. In trade wind belts, 
east coasts are under marine influence and west coasts, under con- 
tinental influence. Oceanic and continental influences on tempera- 
ture are exemplified by the following table of January and July mean 
temperatures across northern North America from west to east: 

Latitude Longitude Mean Temperature F. 

Station North West Jan. July 

Prince Rupert 54 10' 130 6' 35.0 56.0 

Edmonton 53 33' 113 30' 5.5 61.1 

Prince Albert 53 10' 105 38' -4.7 62.7 

Winnipeg 49 53' 97 7' -3.9 66.4 

Ft. Hope 51 33' 87 49' -7.9 62.2 

Moose Factory 51 16' 80 56' -4.8 61.5 

Southwest Point ... 49 23' 63 43' 12.2 56.7 

St. Johns 47 34' 52 42' 23.6 59.3 

The relative importance of continental and oceanic influences on 
the climate of a given region is sometimes expressed numerically 
by an index of continentality based upon the annual range of tem- 
perature as modified by a latitudinal factor. Probably a better meas- 
ure of the relative influence of continents and oceans is the ratio of 
the frequencies of continental and maritime air masses. Unfortu- 
nately, over the greater part of the earth, there are not sufficient 
records to permit the calculation of this ratio. 


Mountain and plateau climates. On mountains and plateaus, the 
average temperature decreases with elevation at a rate approximat- 
ing that of the average lapse rate in the free air, but with many local 
variations. The rainfall increases up to 6,000 or 7,000 feet and then 
decreases because of reduced absolute humidity. As the air becomes 
thinner and freer of dust and moisture, it absorbs less radiation, and 
insolation is, therefore, more intense by day and radiation-cooling 
more rapid by night. However, where there is considerable slope, 
the daily temperature ranges are kept relatively small by the thor- 
ough mixing of the air. 

On large, level plateaus, on the other hand, where there is no aid 
to the mixing of the air, both daily and annual ranges are larger 
than in lowlands similarly situated. In mountain climates, one is 
readily warmed in the sunshine by absorption of the intense insola- 
tion, while the air itself which absorbs little radiation, remains cool. 
At elevations of 12,000 to 15,000 feet, the air becomes so rarefied as 
to cause mountain sickness in many persons because of insufficient 
oxygen. Mountain ranges interfere with the free movement of the 
lower air and often act as climatic divides or barriers, resulting in 
quite different climates on opposite sides. For example, the west por- 
tions of Oregon and Washington have a wet and largely marine 
climate, while east of the Cascade Range the rainfall is light and 
the climate has continental characteristics. The Alps are a barrier 
separating the climate of central Europe from that of the Mediter- 
ranean coast. 

Major climatic factors. Some important influences governing the 
climate of a region and some localities in which each is a prominent 
factor may be listed as follows: 

1. Latitude: Greenland, Amazon Valley, Antarctica. 

2. Position relative to water: Seattle, San Diego, Boston. 

3. Continentality: Oklahoma City, Minneapolis, Verkhoyansk. , 

4. Position relative to mountain barriers: Nevada, Riviera, Ganges 

5. Altitude: Denver, Bogota, Addis Ababa, Tibet. 

6. Prevalence of cyclonic storms: New England, Great Lakes Re- 
gion, Germany. 

7. Prevailing winds: Hawaii, India, Azores, Puerto Rico. 

8. Ocean currents: Norway, Labrador, Northern Chile. 


Climatic Classification 

Irregular distribution and intensity of the climatic factors from 
place to place often renders a latitudinal or topographic zonation of 
climate unreliable and sometimes very misleading. Several classifica- 
tion systems have been devised to overcome this handicap and at 
the same time present the various climatic types in simplified but 
usable form. Each system has its merits, but none is perfect. The 
classification to be reviewed here was first devised by Wladimir 
Koppen in 1918. It has been revised and refined several times. This 
modified version by Trewartha is probably the most popular climatic 
classification in current use in the United States (Fig. 139). It is 

exact formula for^e^ch 

jor~rfiifiatic divisions. Five major climatic regions are recog- 
nized as follows: 

1. A climates, or tropical rainy climates, with no month of the 
year having a mean temperature below 64.4F. (18C.). 

2. B climates, or dry climates, include those arid and semiarid re- 
gions of the world where rainfall deficiency is the prime limiting 
factor in land usability. Evaporation exceeds precipitation, and plow- 
agriculture is not considered feasible without irrigation. The B cli- 
mates are further divided into arid or desert regions (BW) and semi- 
arid or steppe regions (BS). 

3. C climates, or humid mesothermal climates, lie poleward from 
the A climates where the rainfall is sufficient for the area not to be 
classed as B* C climates have a mild winter, the poleward boundary 
being the 32F. (0C.) isotherm for the coldest month. 

4. D climates, or humid inicrothermal climates, lie poleward from 
the C climates and include those regions with long, cold winters but 
adequate summers for agricultural and forestry development. The 
poleward boundary for the D climates is the 50F. ( 10C. ) isotherm 
for the warmest month. 

5. E climates, or polar climates, include those areas having no 
mean monthly temperature above 50F. (10C). They are sub- 

2 Glen T. Trewartha, An Introduction to Climate. New York: McGraw-Hill Book 
Company, 1954. For another recognized classification system, see C. Warren Thorn- 
thwaite, "An Approach Toward a Rational Classification of Climate," Geographical 
Review, Vol. 38 (Jan., 1948), pp. 55-94. 



Fig. 139. Climates of North America. By permission from "An Introduction to 
Climate," 3rd ed., by G. T. Trewartha, 1954. McGraw-Hill Book Company, Inc. 


divided into tundra (ET), with a short growing season, and ice cap 
(EF), where no month of the year has a mean temperature above 

In addition to these five climatic divisions, an H climate, or un- 
differentiated highlands climate, is used to designate those moun- 
tainous areas with topography and exposure so variable within short 
distances as to render the standard classifications impractical. 

Climatic subdivisions. The A, C, and D climatic divisions are 
further refined according to their rainfall and temperature charac- 
teristics. Lower-case letters are used to denote temperature varia- 
tions (not used with A climates) as follows: 

a warm summers (warmest month above 71.6F., or 22C.) 

b cool summers (warmest month below 71.6F., or 22C.) 

c short, cool summers (less than four months above 50F., or 

d-long, cold winters (coldest month below -36.4F,, or -38C.) 
and more specific rainfall distribution data are shown by: 

f no dry season 

s dry summers (rare in the tropics) 

w dry winters, or during the low-sun period 

m dry winters with monsoon rain in summers. 

Thus by knowing the values of the letters representing a climatic 
region, as shown in Fig. 139, it is possible to visualize the climate of 
that region. Geographical position within a climatic region is also 
important, however, and should be considered when the climate of a 
specific place is of concern. For example, the Caf climatic region 
may border the Af to the south, the BS to the west, and the Daf to 
the north. Stations just inside the Caf boundaries on these three 
sides may be radically different from each other. This is not a fault, 
but a virtue, of the system, for the gradual but omnipresent variabil- 
ity of climate creates the need for a simple classification. 

Cyclical Changes of Weather and Climate 

A cycle is the interval of time in which a certain succession of 
events which repeat themselves again and again in the same order, 
is completed. A cyclical, or periodic, event is one that recurs at 
regular, equal intervals. There are two very evident weather cycles 


of such importance that we base our reckoning of time upon them, 
namely, the daily and the annual cycles. It may be noted, however, 
that, as weather periods, neither of these is absolutely regular in its 
recurrence. The diurnal period in the weather is governed by the 
times of sunrise and sunset and is, accordingly, variable in length, 
except at the equator, and altogether ceases to exist as a 24-hour 
cycle in polar regions. The annual period in the weather is also of 
variable length, for the seasons are sometimes "late" and sometimes 

ft f 99 


Weather cycles. A brief examination of a climatic table of rain- 
fall shows that a few dry years often occur in succession, followed 
by a series of wet years, and again by another group of dry years. 
Such short-period fluctuations are constantly occurring, not in pre- 
cipitation records alone, but also in connection with other weather 
elements. It is these variations, and not the daily and yearly periods, 
that are commonly called weather cycles. There can be no doubt 
of the existence of the fluctuations, but none is truly periodic in its 

Much attention has been given to the statistical analysis of 
weather data in the hope of finding periodicities that would be use- 
ful as indicators of future conditions, and a great many so-called 
cycles have been found in this way. There is a list of more than 100 
of these, varying in length from 8 months to 260 years, but they all 
show irregularities: successive recurrences are of different length 
and intensity; the cycles are interrupted by departures in the op- 
posite direction; after persisting for several periods, a cycle may 
suddenly fail, sometimes to begin again later in a different phase. 
There are so many of these "cycles," and they are so irregular, that 
the result can hardly be distinguished from chance. Most of them 
are of doubtful reality. They have not proved of practical value in 
forecasting next year's weather. 

Weather cycles, in the sense of fluctuations of an irregular nature, 
are characteristic of climate throughout the world and often have 
important social and economic consequences. For example, in the 
plains of western Kansas, western Oklahoma, and eastern Colorado, 
there have been series of wet years with accompanying good crops 
and good prices for land, and series of dry years, causing crop fail- 
ures and land abandonment. Cycles of this kind are illustrated in 



Fig. 140, which shows, first, the actual rainfall record for the state 
of Iowa for the years 1873 to 1955, and second, a method of charting 
the data of rainfall to give a clear picture of cyclical changes. 

In the upper curve, the annual total of rainfall is indicated year 
by year by the distance of a point above the zero line of the figure, 
and these points are connected by a broken line. Amounts above 
and below normal occur irregularly, and cycles are not evident in 
this curve. In some cases, a few wet or dry years occur in succession; 
at other times, there are frequent alternations between wet and 
dry years. The wettest year of the record, 1881, is preceded by two 
dry years, and the driest year, 1910, is preceded by two moderately 
wet years. 

In the lower curve, the method of accumulated sums of depar- 
tures is used. The departure of the rainfall of each year from the 
average of the entire series is first obtained, and then, beginning 
with the first year, the year-by-year accumulated algebraic sums of 
these departures are calculated and entered. In a series of wet years, 
there is an accumulating excess of precipitation, and the line moves 
upward; in dry years, it moves downward. The line slopes up when 
the year is wet, and down when it is dry, without reference to its 
position relative to the zero line. 

This curve makes certain cyclical tendencies evident. It shows 
that there was an increasing accumulation of rainfall above normal 
for 11 years from 1874 to 1885, then a declining rainfall with some 
interruptions for 16 years to 1901. There followed a rapid rise for 
8 years with one slight setback. After 1909 the curve is irregular, 
with brief periods of excess and deficiency but with an unmistakable 
downward trend until 1940. The curve again swings upward through 
another irregular, but distinctive cyclic trend. Note that the three 
cycles shown are not of equal length or magnitude. Such imper- 
fectly cyclical variations as are shown by this figure are typical of 
weather records in general and are to be regarded as the natural, 
normal behavior of the climate. Any long-time planning should take 
account of these variations. Unfortunately, planning is hampered by 
the irregular nature of the variations. 

Secular trends. In addition to the short-period variations com- 
monly meant by the term weather cycles, there are tendencies that 
persist over longer periods and are known as secular trends. In Fig. 



141, taken from J. B. Kincer, 3 long-time temperature trends at New 
Haven, Connecticut, and Copenhagen, Denmark, are shown, by 
means of 20-year moving sums. The value entered on the graph 
for each year is the sum of the mean annual temperatures for the 
20 years ending at the year indicated. Thus, the New Haven record 
began in 1780, but the first point on the curve is for the year 1799, 
and the value is the sum of the temperatures from 1780 to 1799, in- 
clusive. The next value, entered at 1800, is the sum for the years 
1781-1800, inclusive, and so on, year by year, for the entire record. 



IMO ItSO IS40 H50 I860 It70 IHO ft0 1000 1910 1*10 1930 




Fig. 141. Secular Trends of Temperature, 1800-1933, Shown by 20-Year 
Moving Temperature Summations. (From ]. B. Kincer.) 

This method of plotting the data largely eliminates the annual fluc- 
tuations and the short-period weather cycles, such as are shown in 
Fig. 103 for Iowa rainfall, and emphasizes the tendencies that per- 
sist for more than 20 years. 

The New Haven curve reaches a peak in 1811 and then trends 
downward until 1875, a period of 64 years, declining rapidly at first 
and then more slowly. From 1875 to the end of the record, the 
general trend is upward, but with some important interruptions 
persisting for a few years. The upward trend had persisted for 57 
years at the close of the record. The curve is below the average 

8 J. B. Kincer, "Is Our Climate Changing?," 

Monthly Weather Review, Vol. 61, 


value continuously for the 58 years from 1820 to 1878, and mostly 
above the average since the latter date. At Copenhagen a sharp de- 
cline in the curve begins in 1837 and continues for 20 years. The 
values then move irregularly upward for a period of 59 years, reach- 
ing a peak in 1915. Since then, the trend has been slightly down- 
ward. The time from maximum to maximum in the Copenhagen 
curve is 79 years. At New Haven the length of the wave, beginning 
with the crest of 1811, was 121 years at the end of 1932, and there 
was then no indication that another crest had been reached. 

Later studies by Kincer to include data to the end of 1945 indi- 
cated that a crest was probably reached at New Haven in 1939. 4 
Kincer found that similar rises, persisting for at least 50 years, had 
occurred generally in central and eastern portions of the United 
States, and that the data for the years 1940 to 1945 suggested the 
beginning of a general reversal of trend. Kincer was premature in 
this assumption, for New Haven's temperatures have averaged 
2.5F. above the normal for the entire period from 1945 to 1955. 
Another evidence of secular trends has been observed in north- 
ern Norway, where winter temperatures have been rising notice- 
ably in recent years, and where glaciers have been decreasing rap- 
idly since about 1890. Such records as these have not disclosed 
cycles of definite length, but the trends are persistent and are large 
enough to modify living conditions in the areas affected. 

There are many evidences of still longer secular trends in climate. 
In studying the annual growth rings of sequoias, A. E. Douglass 
found evidence of oscillations in periods of a few centuries in addi- 
tion to those of a few years. Periods of a similar order of length 
have been found in the study of glaciers and of lake levels in Europe 
and Asia. There is some evidence that Persia and Turkestan, Ari- 
zona and New Mexico, are drier than they were at the beginning 
of the Christian era, and that Yucatan and southern Mexico are 
wetter. If these conclusions are correct, it seems probable tha,t the 
changes indicated are trends or cycles of a still greater length. How- 
ever, in most parts of the world, there is no evidence of important 
trends persisting through centuries. 

When we extend our time scale and think in terms of geological 
epochs, we find that climate has changed greatly, but also cyclically, 

4 J. B. Kincer, "Our Changing Climate," Transactions, American Geophysical Union, 
Vol. 27, No. Ill, June, 1946. 


alternating between glaeial and interglacial periods. At one time, 
glaciers covered large areas of northern United States; at another 
time, some of these areas were covered by dense tropical forests and 
inhabited by huge tropical beasts. Plant seeds and spores preserved 
in peat bogs are now interpreted as giving evidence of several long 
climatic trends since the glaciers disappeared, periods of perhaps a 
thousand dry years and then a thousand wet. The fact that great 
climatic changes have occurred in geologic time is undisputed, but 
the causes of the changes are still a subject of speculation. It is cer- 
tain that alterations in the elevation of the land and in the distribu- 
tion of land and water have resulted in great changes in climate, but 
whether or not there were other causes of the geological climatic 
fluctuations is not known. It seems probable that such slow changes 
in climate are still in progress, as slow changes in the elevation and 
distribution of land undoubtedly are. 

We conclude that there are numerous oscillations in the atmo- 
sphere, some short and some long, and that therefore it is not pos- 
sible to obtain an absolutely stable "normal" value of the weather 
elements. The oscillations resemble cycles but are not truly periodic; 
they resemble the movements of a pendulum, except that the 
weather does not keep time in its vibrations as a pendulum does. No 
physical explanation of the origin and continuance of these oscilla- 
tions is known, They may be the result of variable outside influ- 
ences, particularly insolation, or they may be due to natural periods 
of vibration in the atmosphere itself. They are so numerous, so vari- 
able and inconstant, that thus far it has not been found safe to trust 
their extension into the future. 

Theories of Climatic Changes 

The terrestrial conditions governing the climate of a given por- 
tion of the earth have previously been discussed. In connection with 
the consideration of climatic variability, it is important to examine 
the larger factors that determine the climate of the earth as a whole, 
and whether or not they are subject to slow or sudden changes. The 
four major factors controlling the climate of the world are: (1) the 
output of solar energy: (2) the earth's distance from the sun and 
its position relative to the sun; (3) the extent, composition, and 


dust content of the atmosphere; and (4) the elevation of land and 
the distribution of land and water. 

Solar radiation. As previously noted, the output of solar energy 
has small, irregular variations from clay to day and in an 11-year 
period. There is some evidence, not wholly conclusive, that these 
variations influence weather changes and short climatic fluctuations. 
Other changes in solar energy may have occurred and may still be 
in progress. They may have modified the climate of past ages, but 
no evidence of such changes in solar output exists. In particular, 
there is no reason to suppose that sudden changes of climatic sig- 
nificance have occurred. 

Relative position of earth and sun. Aside from the regular sea- 
sonal variations in the position of the earth's axis relative to the sun 
and in the earth's distance from the sun, these undergo slow and 
slight changes in periods of from 21,000 to 400,000 years. One theory 
of the cause of glacial and interglacial epochs, Crolls theory, is 
based upon these recurring slight changes in the earth's orbit. The 
theory is open to serious objection as an explanation of the known 
glacial history, and, in any case, such changes as have occurred 
within the past few thousand years have had no appreciable effect 
on climate. 

Atmospheric content. The extent and gaseous composition of 
the air, by affecting the amount of absorption of incoming or out- 
going radiation, affect the climate of the world. Another theory of 
the cause of the ice ages is based on supposed variable amounts of 
carbon dioxide in the air in different eras, and the fact that this gas 
is a good absorber of earth radiation. This theory is not generally 
accepted, and it appears certain that the proportion of the gases of 
the air has remained practically constant since the beginning of his- 
tory, although it probably changed appreciably in geological epochs. 

During past geologic ages, there were probably periods of great 
volcanic activity during which immense quantities of dust were 
thrown into the air. This volcanic dust, by intercepting much solar 
radiation, may have been an important factor in the production of 
climatic changes and, according to Humphreys, was probably one 
of the chief causes of glaciation. There is observational evidence 
that large volcanic eruptions in historical times, such as those of 
Krakatoa in 1883 and Katmai in 1912, were followed by cooler 


weather for a year or two. These slight temporary results have been 
observed, but the variation in the amount of volcanic activity in the 
past few thousand years has not been sufficient to effect a persistent 
change in climate. 

Distribution and elevation of land. Finally, great changes in the 
elevation of large land areas, in the extent of the land surface, and 
in the distribution of land and water have undoubtedly caused great 
alterations in the climates of the world in the past million years. 
There are evidences of the alternate uplift and subsidence of large 
land masses, resulting in great variations in the elevation of the land 
and also in the ratio of the total land surface of the globe to the 
water surface. We know that elevation has important effects on cli- 
mate, and we know that changes in the extent and position of land 
areas would greatly modify climate, not only because of the differ- 
ent responses of land and water to insolation, but also indirectly by 
producing changes in ocean currents and atmospheric circulation. 
It seems clear then that these terrestrial changes have been impor- 
tant factors in past climatic pulsations, but the changes are slow in 
terms of man's history, and such slight changes as have occurred in 
historical times have had no observable effect. 

Stability of historic climate. We may disregard geological 
epochs, because they are too long to be included in the ordinary, 
everyday meaning of climate, and we may disregard weather cycles 
as too short, and may consider climate to mean the summation of 
weather conditions within the recorded life of man. In this sense, 
climate is about as stable as anything we know on earth, about as 
permanent as the hills. While there is some evidence in Asia and in 
our southwest of changes in the past 1,000 or 2,000 years, in most 
parts of the world the evidence is to the contrary. Olives are still 
grown in Palestine and silkworms in China, under apparently the 
same climatic conditions as prevailed several thousand years ago. 
In spite of weather cycles and secular trends, the climates of the 
world appear not to have changed progressively in one direction 
within the period of history. 

There are no sudden, violent changes of climate. That is the con- 
clusion to be drawn from our knowledge of the past, and it is also 
the conclusion when we consider the causes of climate, that is, the 
climatic controls discussed in the preceding paragraphs. While all 
of these factors are more or less variable in the slow course of time 


and may have been influential in producing geological changes of 
climate, we have no reason to suppose that any of them ever has 
or ever will change suddenly or appreciably within a few hundred 
or a few thousand years. We may therefore expect the climates of 
the world to remain relatively stable in terms of human history. 

Further, it is evident that the activities of man cannot influence 
these major controls of climate. We cannot yet analyze all the forces 
affecting weather and climate nor explain their periodic fluctuations, 
but both reason and experience indicate that climate is much more 
stable than human institutions or relations. The climatic factors, 
affecting profoundly the economic, social, and physical life of man, 
remain comparatively permanent in a changing world. Nations rise 
and fall, causing changes in trade routes, the rise of new commercial 
cities, and the decline of old ones. Scientific discoveries and their 
applications lead to new industries and new habits with resulting 
changes in economic life and the distribution of the population. 
Climate, however, remains a practically constant element of man's 


1. On three outline maps of the world, draw the isotherms for the year, 
for January, and for July. 

2. On each of these maps, draw the "heat equator" through the middle 
of the hottest belt. 

3. Why do higher maximum temperatures occur in North Dakota than 
in Alabama? 

4. Indicate the normal annual precipitation on a map of the world by 
different colors or shades for intervals of 10 inches. 

5. Why do continental interiors have their maximum precipitation in 

6. Obtain a local Annual Meteorological Summary as published at 
many Weather Bureau stations, and list the various items of climatid in- 
formation published therein. 

7. Plot records of temperature and rainfall, using accumulated sums 
of departures and also 20-year moving sums, and report on the charac- 
ter of the variations indicated. 

8. Draw a five-inch square to represent North America, and schemati- 
cally subdivide it to show the climatic classification according to Fig. 132. 

9. Is the climate of your state changing? 





A study of the general and secondary circulations makes it evi- 
dent that the atmosphere is fluid and mobile and acts as a whole. 
Anything that happens in one part of it affects it all. Local weather 
and climate are small portions of world weather and climate. The 
characteristics of weather and climate, and the way in which they 
vary in time and place, have many important relations to the life 
and labors of man. In fact, they are fundamental factors condition- 
ing our life on this planet. In this chapter attention is called, briefly, 
to some features of the weather in its world-wide relationships and 
to some phases of man's response to his climatic environment. 

Variability of the Weather 

The weather vane has long been a symbol of fickleness. Change 
and variety are characteristic of the weather outside of the tropics, 
in contrast to the monotony that often prevails in trade-wind and 
equatorial climates. 

Combinations of weather elements. The weather elements, such 
as temperature, precipitation, wind direction and speed, humidity, 
sunshine, and cloudiness, are all continuously variable within rather 
wide limits, and to a certain degree independent of one another. 
Hence, the number of possible combinations among them is very 
great. When 1,440 minutes of changing atmospheric conditions are 
combined to make a day of 24 hours, the number of possible permu- 
tations of the weather elements becomes almost infinite, and we see 
why no two days, as no two human faces, are exactly alike. 



It is evident at once, however, that we do not experience our 
weather entirely by a random sampling of the possible combina- 
tions of the weather elements. There is, for example, the seasonal 
control of temperature; in the United States we do not have zero 
temperatures in July nor 100 in January. It is also clear that the 
different meteorological elements do not vary with complete inde- 
pendence. There is an evident correlation between wind direction 
and temperature, between wind direction and rainfall, and between 
sunshine, cloudiness, and rainfall. The number of combinations is 
considerably restricted by these relations and further limited by the 
fact that the weather of a given day is not completely independent 
of the preceding day's weather, as will be noted in the next section. 
In spite of these limitations, each day is unlike every other, and 
sometimes the changes from one day to the next are extreme. At 
Goodland, Kansas, on January 26-27, 1951, the temperature dropped 
from 79F., to 3F. in 18 hours. This is an ^extreme case, having oc- 
curred during one of the most severe cold waves in the history of 
the Weather Bureau. Wide variations in temperature from day to 
day are frequent in the winter months, however, throughout most 
of the United States. Rapid and large falls are more frequent than 
similar rises. 

Monthly and annual variability. When these erratic days are 
combined into weeks and months, and the months into seasons and 
years, we get an immense number of possible groupings and an in- 
finite variability of detail. Months whose average conditions are the 
same may in fact be very different, when compared, from day to 
day. The average temperature of the month of February, 1933, at 
Des Moines, Iowa, was nearly normal, but half the month was ex- 
tremely cold, and the other half was unusually warm. There were 
no normal days, but the month was normal! Rainfall may be equally 
erratic. The distribution of the various amounts of precipitation in 
January at Cleveland, Ohio, for 64 years, is illustrated in Fig. 142, 
which is called a frequency polygon. It will be noted that the figure 
is one-sided; there are more small values than large ones. The aver- 
age, or mean, value is 2.54 inches, but the most frequent value, or 
mode, is from 1.40 to 1.59 inches, and 2.23 is the median, or middle 
value, when the amounts are arranged in the order of their magni- 
tude. A longer record might alter this distribution considerably. 

Rainfall usually has an unsymmetrical distribution, often more so 



than is shown in this case, especially in drier regions. In parts of the 
semiarid west, a rainfall of 12 inches in a year is considered suf- 
ficient for the growing of dry-land grains, but in that region an an- 

Fig. 142. Frequency Polygon of January Precipitation at Cleveland, Ohio, during a 
64-Year Period. Mean, 2.54 inches; median, 2.23 inches; mode, 1.40 to 1.59 inches. 

nual average of 12 inches is usually made up of a few years of much 
more than the average and many years with amounts somewhat 
less than average. The farmer using the land under these conditions 
should realize that the land will receive less than 12 inches of pre- 
cipitation more than half of the time, 

In general, temperature data are arranged according to chance 
and, when plotted, form a symmetrical curve in which the mean, 
median, and mode coincide. This is illustrated in the polygon and 
curve showing the variations in the length of the growing season at 
Indianapolis, Indiana, Fig. 143. Frequency curves may be drawn by 
inspection as in this figure; but if a more accurate representation of 
the data is required, the algebraic equation of the curve may be 
calculated. Mathematical considerations also enable us to determine, 
within a margin of error dependent upon the length and character 
of the record, how often the rainfall will fall below or exceed any 
given amount, or to determine the probability of a killing frost after 
a given date in spring or before a given date in autumn. 

By such mathematical means it becomes possible from the exami- 



nation of a limited number of observations to obtain a reasonable 
estimate of events as they will occur in the future on the average, a 
more accurate estimate than can be obtained by simply counting 
the number of times the given events have occurred in the past. Of 
course, it is not possible in this way, nor in any other way now 
known, to accurately predict when the favorable or unfavorable 
seasons will occur. They may appear to happen fortuitously, but in 
the long run they will occur the number of times indicated by the 

Fig. 143. Frequency Polygon for Length of Growing Season at Indianapolis, In- 
diana, for a 60- Year Period and Curve Showing the Probable Distribution of 
Frequencies in a Very Long Record. The curve is approximately symmetrical about 
the mean value. Mean, 188 days; median, 189 days; mode, 190-199 days. 

curve, and it is the performance of the weather and the yield of the 
land in the long run that determine values, although to the indi- 
vidual owner the events of a few specific years may be of first im- 

Persistence of the Weather 

In contrast with the character of changeableness which we asso- 
ciate with atmospheric phenomena, we sometimes find the weather 


in a less fickle mood. There are times when similar weather condi- 
tions continue day after day for considerable periods. The per- 
sistence of the weather is illustrated by what are called weather 
"spells" and weather types. 

Weather spells. Such phenomena as the occurrence of rain on 
several successive days and the persistence of hot weather for a 
week or two are familiar, and are familiarly called rainy spells and 
hot spells. When one kind of weather is established, it sometimes 
has a tendency to continue for several days; if it rained yesterday, 
the chances of rain today are better than if it was fair yesterday. 
Today's weather is not independent of yesterday's. As mentioned in 
the previous section, there are times of rapid variability of the 
weather, but, on the average, similar weather tends to persist for 
several days. Reasons for this persistence may be found in the in- 
fluence of the semipermanent areas of high and low pressure of the 
general circulation, and in the slow movement or stagnation of 
cyclones and anticyclones, resulting in the continued inflow of warm 
air or outflow of cold air. 

Weather types. Further examination of the records discloses per- 
sistence of another kind. We find periods in which the abnormal 
conditions are not absolutely continuous, day after day, but in which 
the same kind of weather recurs frequently. It is evident that 
months of unusual departures, having very heavy rain, for example, 
or averaging markedly cold or hot, indicate the continuance of ab- 
normal conditions for at least the greater part of a month. Do such 
departures continue for more than a month? Fig. 144 shows the 
deviation of the mean monthly temperatures from the average at 
St. Paul, Minnesota, for each month in succession from January, 
1929, to December, 1955, inclusive. 

It will be seen that there are frequent alternations above and be- 
low normal, but these do not appear to be systematic; no law of 
variation is evident. Note that January, 1944, was a very warm 
month between two cold ones; from February, 1933, to December, 
1935, there were fairly regular monthly variations between positive 
and negative departures, and February, 1936, was a very cold 
month separating two normal months. Yet, more often than not, 
we find two or more months in succession on the same side of the 
normal. Each of the first 6 months of 1952 was warmer than normal, 
and the 6 months from March, 1940, to August, 1940, were all colder 




than normal. Note especially the long "warm spells" in 1930 and 
1931, when there were 9 consecutive months with positive depar- 
tures, and again in 1941 and 1942, when 13 consecutive months 
averaged 4.5F. above the normal. In 1948 and 1949, there was a 
period of 7 warm months, and there are a number of other periods 
in this record with from 5 to 6 consecutive months having depar- 
tures of the same sign. 

These records illustrate the tendency for similar temperature de- 
partures, that is, similar types of weather, to continue for several 
months, but, as has been seen, there are many exceptions. Rainfall 
curves show variations of the same kind. There are dry seasons with 
the rainfall deficient for several months in succession, and, at other 
times, wet periods of a few months' duration. 

These are examples of the fact, with which students of weather 
in temperate latitudes are familiar, that similar weather conditions 
often persist for periods varying from a week or two to several 
months, and then change abruptly to weather of quite a different 
character. As has been noted in Chapter 12, upper-air pressure 
contour charts give some indication, for short periods in advance, of 
the tendency toward persistence or toward change. The persistence 
of a given type of weather probably means that the general pres- 
sure distribution remains approximately the same, and that depres- 
sions or anticyclones of like characteristics follow one another in 
succession along about the same paths. 

In a warm winter in the United States, for example, many lows 
move across the northern border, and few cold air masses push 
southward from Canada. In a wet season in the Mississippi Valley, 
many depressions originate in the southwest and travel northeast, 
bringing much moist air from the Gulf of Mexico. The change to an- 
other type of weather occurs when the highs and lows take a dif- 
ferent course. The reason for their taking a different course is to be 
found in some alteration in the general circulation. In the Northern 
Hemisphere, such changes in the general circulation are usually 
shown by modifications in the position and intensity of those sea- 
sonal centers of action, the Aleutian and Iceland lows, the Bermuda 
and Pacific highs, and the continental highs of winter over Canada 
and Siberia. Modifications in the west-to-east circulation are ex- 
pressed in terms of the circulation index and appear in the circula- 
tion pattern as shown on upper-air charts. 


Weather Correlations 

Such changes in the general circulation as have been mentioned 
in the preceding section affect the whole atmosphere and conse- 
quently result in weather changes throughout the world, but not 
necessarily changes of the same kind. An alteration in the paths of 
traveling disturbances may bring unusually wet weather to one re- 
gion and dry weather to another, or cold air to some areas and warm 
to others. Furthermore, the response to pressure changes is not im- 
mediate; there is a time lag between cause and effect. A rise of pres- 
sure in one part of the world may show itself several months later 
in changed weather conditions in a distant part of the globe, not 
necessarily in the same hemisphere. 

Correlation coefficients. A correlation coefficient is a numerical 
quantity that expresses the degree of linear 'relationship or corre- 
spondence between two sets of data. It is computed by a mathe- 
matical process which takes account of and compares the individual 
deviations from average of the two sets of data. For example, when 
the average pressure during the summer months at Honolulu is com- 
pared for a series of years with the average temperature during the 
following winter months in the Missouri and upper Mississippi Val- 
leys, a certain degree of correlation is found to exist. It indicates 
a tendency for high summer pressures at Honolulu to be followed 
by cold winters in the Missouri and upper Mississippi Valleys, and 
low pressures by warm winters, but the relationship is not close 
enough to be of forecasting value. 

There is a similar relation between the pressure in South Amer- 
ica and subsequent rainfall in India and temperature in Japan. If 
the pressure is unusually high in Argentina and Chile during March 
and April, there is likely to be a heavy monsoon rainfall in India in 
the following July and August, and a wkrmer-than-normal August 
in Japan. There is a negative correlation between summer rainfall 
in Cuba and the next winter's rainfall in southern England; a rela- 
tion between the spring temperatures in Siberia and the summer 
temperatures in California. 

Many other correlation coefficients have been obtained, showing 
the existence of similar correspondences between widely separated 
areas* There is a well-known negative correlation between the pres- 


sure over Iceland and that over the Azores. When the Iceland low 
is unusually deep, the Azores high is also strongly developed; like- 
wise, when the low is shallow, the high is weak. A similar oscilla- 
tion occurs in the north Pacific Ocean between the Aleutian low 
and the Pacific high. In the Southern Hemisphere, there is a nega- 
tive correlation of pressure between the south Pacific and Indian 
oceans. It is evident that these relations between distant weather 
conditions are expressions of the unity of the air. The general dis- 
tribution of pressure and the general circulation of the air undergo 
changes that are reflected in world-wide modifications of weather. 1 

The Oceans and the Weather 

The study of ocean temperatures and ocean movements in rela- 
tion to atmospheric changes gives evidence of their interdepend- 
ence and of the world-wide relations of weather phenomena. For 
many years, records of water temperatures at the surface and rec- 
ords of weather conditions have been obtained by ships as they 
travel their regular routes. A more intensive program was conducted 
during World War II, and much data were compiled concerning 
currents, water temperatures at various levels, and surf conditions, 
and their relations to wind and weather. Thus our knowledge of the 
behavior of the air and our ability to foresee its changes have been 
enlarged, but lack of data from the oceans still limits our knowl- 
edge of the complex relations between the oceans and the weather. 
A fully adequate scheme of observations would require a great num- 
ber of continuous records from fixed positions in all the oceans. 

Some relations between ocean temperature and weather. Be- 
cause water changes its temperature very slowly, the ocean waters 
are great conservers of heat, and by their movements they are great 
transporters of heat and equalizers of temperature. The heat carried 
by ocean currents from the tropical waters of America to the north 
Atlantic saves the people of northern Europe many thousands of 
tons of coal each winter, for some of that heat is transported to the 

1 For further discussion of statistical methods applied to meteorological and climato- 
logical data, see C, F. Marvin, "Elementary Notes on Least Squares, the Theory of 
Statistics and Correlation for Meteorology and Agriculture," Monthly Weather Re- 
view, Vol. 44 (1915), pp. 551-569, and V. Conrad and L. W. Pollak, Methods in 
Climatology, 2nd ed., Cambridge: Harvard University Press, 1950. 


land by the winds which are warmed in their passage over the warm 

Changes in ocean temperature not only effect changes in the tem- 
perature of the land to leeward, but they produce other, less direct, 
effects. Temperature variations over large areas result in a redis- 
tribution of atmospheric pressure with varied and far-reaching influ- 
ences on weather conditions. For instance, the presence of unusually 
warm water off the southeastern coast of the United States during 
the winter months probably results in colder-than-normal weather 
in the eastern states, instead of warmer, as might at first be assumed. 
A large body of warm water tends to reduce the pressure in its 
vicinity, and in this case would intensify the pressure gradient be- 
tween the ocean and the winter high-pressure area over the north- 
ern continental interior. Hence, the eastern states would receive 
more than the usual amount of cold air from interior Canada. In 
the following paragraphs some specific illustrations of the interrela- 
tions of air and water conditions are given. 

Effects of northeast trades. The changes in the temperature of 
the ocean water at any one place are not due greatly to the heat of 
the sun there nor to the temperature of the air over the water. They 
are due chiefly to the effect of the winds in moving the water. For 
example, strong, steady, northeast trade winds in the north Atlantic 
cause the warm surface water to drift toward the West Indies and 
the Caribbean Sea, resulting in an accumulation of warm water in 
those regions. The place of the water thus removed is taken either 
by colder water formerly beneath the surface or by colder water 
drifting in from more northerly regions. Thus, the effect of unusu- 
ally strong trades is to make the water colder in the mid-Atlantic 
trade wind area and warmer in the Caribbean and the Gulf of 

A portion of this warm water moves through the Straits of Florida 
and thence along our eastern Atlantic coast and across the Atlantic 
toward the British Isles and Scandinavia. The amount of water thus 
transported is probably more than a thousand times the average 
discharge of the Mississippi River. Since it takes 3,000 times as much 
heat to warm a given volume of sea water by 1 as to warm an equal 
volume of air by an equal amount, the effect of such a volume of 
water on air temperatures is great. It is believed that enough warm 


water is carried by the North Atlantic Current into the Norwegian 
Sea each year to raise the temperature of the air over the whole of 
Europe up to 2% miles above the surface of the earth by 10 de- 
grees for each degree that the water cools. 

Polar ice and the weather. The surface area of floating ice in the 
polar seas not only undergoes a seasonal change but frequently 
shows large variations from one year to another. An increase in the 
amount of ice is attended by a decrease in the temperature of the air 
over the ice and over adjacent regions and a related increase in 
pressure over these regions. In the north Atlantic, when the ice in- 
creases, the general tendency is toward a filling up of the Iceland 
low and a flattening of the Azores high. These conditions result in 
altering the paths of traveling depressions across the Atlantic. In 
particular, changes in the amount of ice in the Greenland Sea region 
are thought to be an appreciable factor in variations in the weather 
of the British Isles and Norway. 

The Peru current and Peruvian rainfall. A striking example of 
the effect of ocean changes upon the weather of near-by land areas 
occurred on the Peruvian coast of South America from January to 
April, 1925. Ordinarily the cold Peru current from the south pre- 
vails along those shores, somewhat mitigating the heat of the ad- 
jacent lands but causing them to be almost rainless, because the 
cool air is warmed as it moves inland and its relative humidity 
thereby reduced. During the early months of 1925, this current 
seems to have disappeared and to have been replaced by a warm 
northerly current from which warm, moist air moved inland. 

The reason for this departure of the ocean from its well-estab- 
lished habit is not known, but it was doubtless meteorological in 
its nature, caused by some variation from normal temperature and 
pressure somewhere in the world. The climatic consequences were 
remarkable. In desert regions where rain was almost unknown, great 
floods spread destruction and dismay. Counterbalancing these losses 
came a quick and abundant growth of grass, giving the half -starved 
animals such a feast as they had not known for years. During April, 
conditions returned to normal. Such experiences occur at irregular 
intervals of several years. 

The fact is, then, that ocean circulation and ocean temperature 
are closely connected with air circulation and air temperature. 
Changes in either cause changes in the other, sometimes in distant 


parts of the world and after the lapse of considerable time. The vari- 
ous influences interact inextricably, but it seems clear that a more 
complete knowledge of ocean currents and ocean temperatures and 
their variations from season to season and year to year, together 
with a broader knowledge of atmospheric changes over the oceans, 
would be of value in interpreting what often appears to be the 
capricious behavior of the atmosphere. 2 

The Sun and the Weather 

The sun not only governs the movement of the world in its orbit, 
but also, by its never-ending stream of radiant energy, it rules the 
earth's life and activity. The sun's general control of the earth's 
weather and climate is evident, but how detailed its regulation of 
the weather is, remains a question. As previously noted, there are 
slight variations in the flow of energy from the sun; the solar con- 
stant appears to vary slightly from day to day, and in longer periods 
somewhat more, up to about 3 per cent. Are these solar changes 
reflected in weather changes on the earth? 

Sunspots and weather. The number of sunspots increases and 
diminishes in a cycle averaging about eleven years, from one maxi- 
mum or minimum to the next, attended by changes in the solar out- 
put of radiant energy. Evidences of this period appear in some 
weather records and in the thickness of tree rings and of clay layers 
deposited by glaciers. 3 Records indicate that atmospheric pressure 
is relatively low in the tropics and high in high latitudes during 
periods of great sunspot activity, and the reverse during periods of 
very little sunspot activity. Sunspot periods are of varying length 
and intensity, and the weather responses are changeable in amount, 
difficult to follow, and often obscured by other fluctuations. 

Solar radiation and the turbulent atmosphere. A change in the 
intensity of solar radiation may cause a shifting of the pressure belts, 
with consequent complex effects on the paths of cyclones and anti- 
cyclones and upon the distribution of temperature and precipitation. 
It is reasonable to assume that the daily fluctuations tod the slow 
cyclical variations in the output of radiation by the sun affect the 

2 For further discussion of ocean and weather relations, see Sverdrup, Oceanog- 
raphy for Meteorologists. Englewood Cliffs, N. J.: Prentice-Hall, Inc., 1942. 

3 Harlan T. Stetson, Sunspots in Action. New York: The Ronald Press Company, 
1947, pp. 160-177. 


weather throughout the world by affecting the temperature of the 
earth and the air, but just how and to what extent they may influ- 
ence or control the observed weather changes are questions not yet 
satisfactorily answered. Our erratic day-to-day weather appears to 
result from innumerable differences in the physical condition of the 
air. Even with a constant amount of heat from the sun, the physical 
condition of the lower air would be subject to countless local 
changes because of such factors as the variations in the surface cov- 
ering of the earth and differences in elevation, absorption, radiation, 
evaporation, cloudiness, and dust. The influences affecting the air 
appear to be so numerous, so immeasurable, and so unpredictable as 
to create a condition of extreme atmospheric turbulence defying 
exact analysis and never twice the same in detail. 

Seasonal Forecasting 

The object of the preceding discussions of weather correlations 
and solar and oceanic influences has been not so much to give def- 
inite results of immediate application as to emphasize the com- 
plexity of world weather relations. The facts now known as to these 
relations give us some indication of how the air behaves, but our 
knowledge is not sufficiently definite or complete to enable us to 
explain fully the physics of the air in its larger movements. Attempts 
have been made to apply the existing knowledge of the world-wide 
relations of weather to the problem of long-range forecasting mean- 
ing the forecasting of the general character of the weather for a 
month or more in advance. An attempt is made not to forecast the 
daily weather, but rather to say whether the precipitation and the 
average temperature of the period under consideration will be above 
or below normal, and how much above or below. Seasonal forecast- 
ing is the attempt to foresee the general character of a future sea- 
son, for example, to determine whether the coming winter will be 
warmer or colder than normal. Such attempts are usually made for 
seasons not more than six to nine months in the future. 

Nature of atmospheric responses. It sometimes appears that those 
atmospheric responses to changing conditions that show themselves 
in different parts of the world are to be explained as latitudinal shift- 
ings of the pressure belts as a whole or in large areas, such as it has 


below freezing; on the other hand, such tropical plants as date palms 
require a temperature of about 64F. to start growth. The growth 
of citrus fruits is limited by temperature to small areas in the United 
States. Often small, local differences of temperature determine the 
selection of land for oranges in southern California. Outside of 
the tropics, most agricultural crops begin to grow at about 43F. 
(6C.), but growth is most vigorous and healthy when the tempera- 
ture of the soil is between 65 and 70F. Perennial plants retire into 
a rest period when the temperature is below 43. 

Precipitation is essential to supply the moisture by which food is 
taken from the soil in solution and carried throughout the plant by 
the sap. It is also necessary to prevent the drying and wilting of the 
leaves, from which large quantities of water are transpired to the air 
in the growth processes. The significance of precipitation as an asset 
is illustrated by the difference between eastern Kansas and western 
Kansas in the value of land and the density of population. 

Weather and crops. Climate largely determines what shall be 
the staple crops of a region; the weather of the individual seasons 
largely determines the yields of those crops. It is often not so much 
the total rainfall and the average temperature that fixes the yield, as 
the distribution of moisture and favorable or unfavorable tempera- 
ture through the season. There are certain short critical periods in 
the growth of many crops during which their success or failure is 
largely determined. With some crops and in some climates, tem- 
perature is the controlling factor; with others, it is rainfall or sun- 

For example, corn can recover from earlier droughts, but if it suf- 
fers for moisture while tasseling, the yield will be small. J. Warren 
Smith found that the first ten days of August are the most critical 
in the production of corn in Ohio. This is the time when a good 
shower is truly a "million-dollar rain." Winter wheat needs cool and 
moist weather while growing rapidly, but warm and dry weather 
while the heads are forming and filling. Potatoes require cool 
weather with plenty of moisture, especially during the ten days 
following blossoming. The rainfall of May largely influences the 
production of hay in a large part of the United States. A relatively 
cool and wet August is of importance in the production of cotton 
in our southern states. 

The relations just mentioned and many similar ones have been 


discovered largely by the methods of correlation. Records of crop 
yields are compared year by year with records of temperature, rain- 
fall, and other weather elements, and the influences of the weather's 
variations on the yield are determined. The knowledge of these in- 
fluences helps in the adjustment of crops to the most favorable cli- 
matic conditions. It helps to decide whether a given crop is well 
adapted to a given region. Sometimes by the use of different va- 
rieties or different methods of cultivation or by varying the time of 
seeding, the time of occurrence of the critical periods can be ad- 
justed to the time when the weather is most likely to be favorable. 
It is obvious that a knowledge of the water requirements of plants 
at different stages of their growth is particularly applicable to farm- 
ing under irrigation. 

The factors influencing the yield of a given crop at a given time 
and place are many and complex, and include condition of soil, 
tillage, and seed, in addition to the weather. The efficiency of a 
given amount of rain in producing a crop is influenced, among other 
things, by the amount of evaporation. In the arid southwest, evap- 
oration is great, partly because of low humidity and the infrequency 
of clouds, together with high average temperatures, and crops re- 
quire more water there than they do in cooler, cloudier regions. In 
addition to the direct effects of atmospheric conditions on yields, 
there are the indirect effects resulting from the development of plant 
diseases and insect pests. Grain rusts are encouraged by hot and 
humid weather at the ripening stage of the grain; the extent of boll 
weevil damage to the cotton crop is related to sunshine and humid- 
ity during the growing season, and also, since low temperatures kill 
the weevil, to the minimum temperatures of the preceding winter. 

Droughts. A drought is a continued lack of moisture, so serious 
that crops fail to develop and mature properly. The dry period is 
of particular significance when it is of unusual length as compared 
with normal conditions in the area, A period of two summer months 
without rain would not be serious in California, because it is the 
usual thing, but one rainless summer month in central and eastern 
portions of the country would constitute a severe drought. The 
severity and effect of long dry periods depend not only on their 
duration but also on the attending temperature and wind, on the 
kind and previous condition of the soil, and on the condition of 


the crops. For these reasons, no exact definition of a drought in terms 
of the number of rainless days can well be given. 

Phenology. In a great degree, the growth of plants is a response 
to the atmospheric influences to which the plants have been sub- 
jected, a summation of weather influences. Plants are nature's record 
of the climate and weather. A record of the progress of plants 
through the growing season is, then, in part, a weather record, A 
record of the time of leafing, blossoming, and fruiting gives an in- 
dication of the progress of the seasons. Averages obtained from 
these records are a function of the climate of a region, and the yearly 
variations from the average are an indication of how the weather 
has varied. These data form the basis of the science of phenology, 
which may be defined as the study of the phenomena of life, espe- 
cially plant life, as they recur from year to year, and their responses 
to weather and climate. Although agricultural practices are in the 
main determined by long experience, the knowledge gained through 
phenological records aids in the adjustment of farming operations 
to the most favorable weather conditions. The data are indispen- 
sable in the calculation of correlations between weather and crop 


1. Draw frequency polygons and frequency curves from monthly or 
annual temperature and precipitation tables. 

2. Make graphs of successive departures from normal of monthly tem- 
peratures and rainfall and note the character of the curves with refer- 
ence to variability and persistence. 

3. Chart daily records of temperature and rainfall for evidences of 
short weather "spells." 

4. Examine records over several months for evidences of weather types. 

5. By spending a month in eacri of twelve cities in the United States, 
one might pass the year under nearly ideal temperature conditions, 
assuming that normal temperatures prevailed. What cities might one 

6. Discuss what period without important rain constitutes a drought 
in your state. How is the length of the period related to the time of year, 
temperature, wind, and previous condition of the soil? 

7. What amount of rainfall is required to break the drought? How 
frequently do such droughts occur? 




Meteorological services require a widespread, systematic network 
of stations for the observation, collection, and dissemination of 
weather data. They require an organization covering an entire na- 
tion and must interchange data freely with other nations. Reliable 
weather forecasts are essential to many activities and desired by 
everyone; they should therefore be unprejudiced, impartial, and 
freely accessible to all. 

In addition to the public services, a private individual, company, 
or other organization may employ a meteorologist to provide special 
forecasts or other services to meet the needs of the employer. Thus, 
citrus fruit growers, commercial airlines, oil companies, shippers, 
and the armed forces frequently provide meteorological consultants 
or forecasters to meet their particular needs which are not com- 
pletely satisfied by the public forecasts. 

Brief History of Meteorology 

The science of weather was born in the twilight of civilization, 
but remained in its infancy for countless centuries owing to the lack 
of instrumentation or of a knowledge of atmospheric dynamics. 
Principles of physics, mathematics, chemistry, geography, astron- 
omy, and mechanics had to be discovered before meteorology could 
become a true science. The study of weather was aided by Torri- 
celli's barometer in 1643, Fahrenheit's thermometer in 1710, and 
several other measuring instruments which came in rapid succession. 
It was gradually developed by such men as Copernicus, Kepler, 
Newton, da Vinci, Galileo, Halley, Dalton, Charles, Boyle, Gay- 



Lussac, Buys-Ballot, Ferrell, and many others who made significant 
contributions in their particular scientific fields. 

Hippocrates wrote a treatise on medical climatology about 400 
B.C., and fifty years later Aristotle wrote a book based on his per- 
sonal weather observations. The real turning point in the maturity 
of meteorology as a science did not occur, however, until the early 
part of the twentieth century, when V. Bjerknes developed his air 
mass and wave-cyclone theory of weather propagation, and the de- 
velopment of aviation created an urgent need for more accurate in- 
formation about the weather. 

Meteorology has not yet reached the point of becoming an exact 
science, but recent theoretical developments, improved instrumenta- 
tion, and electronic computers will surely permit weather forecast- 
ing to become more and more reliable with time. 

Development of a Weather, Service 

In the early history of this country, only a few individual weather 
records were recorded over extended periods. Some of these have 
been preserved and are of much value, but there was no systematic, 
organized collection of data. In the rapid development of the coun- 
try during the first half of the nineteenth century, the need for in- 
formation concerning weather and climatic conditions was widely 
felt, and numerous governmental agencies began to collect such 
data as an incidental and rather extraneous part of their work. 

Early organized weather observations. The first of these obser- 
vations was made by the Government Land Office, which began in 
1817 a system of precipitation records and tri-daily observations of 
temperature at its various local offices, which were widely dis- 
tributed in the newer portions of the country. In 1819, regular ob- 
servations were begun at the military posts throughout the country. 
In 1841, the Patent Office organized a body of weather correspond- 
ents, or voluntary observers, and inaugurated systematic observa- 
tions. The Smithsonian Institution did likewise in 1849, and in 1857 
it began receiving telegraphic reports of simultaneous observations, 
from which Professor Joseph Henry, then Secretary of the Smith- 
sonian Institution, prepared weather maps and forecasts. This serv- 
ice ceased in 1861 owing to the outbreak of the Civil War, 

After the Civil War, Professor Cleveland Abbe at Cincinnati, 


aided by the co-operation of business organizations and the tele- 
graph company, arranged to receive telegraphic reports of weather 
conditions in 1869 and 1870. From these he prepared daily synoptic 
charts and issued statements of weather "probabilities," especially 
with reference to storms on the Great Lakes. The time was now ripe 
for the organization of a government weather service in the United 
States, as had already taken place in some European countries, and 
Increase A, Lapham of Milwaukee was instrumental in having in- 
troduced into Congress a bill providing for such an organization. 
This bill became a law on February 9, 1870. The law required that 
the Signal Service of the United States Army make meteorological 
observations at its military stations and at other points, and that it 
give notice of the approach and force of storms. General Albert J. 
Myer, Chief Signal Officer, proceeded to the work of organizing the 
new service, which began operations on November 8, 1870, with 
Lapham as meteorologist. In 1871 Abbe also became connected with 
the official weather service and continued as one of its leading sci- 
entific officials until his death in 1916. 

The official weather service continued to be conducted by the 
Signal Service for the next twenty years, with gradual extension of 
observational stations and increasing usefulness to the public. Al- 
though in the beginning the service was intended primarily for the 
issuance of storm warnings for the benefit of navigation, it soon 
became evident that valuable information could be given to a much 
wider public. The service could be useful not only in the matter 
of forecasts, but also in disseminating the facts of existing weather 
conditions and of the climatic characteristics of the country. Thus 
there came into existence in response to public demand the three 
main features of a weather service, namely, the accumulation and 
publication of climatological data, the preparation and distribution 
of weather forecasts, and the dissemination of current weather in- 

Establishment of the Weather Bureau. During this period of 
growth, it was necessary to develop the technique of observation 
and forecasting, to devise and improve instruments, to train and in- 
struct many observers, and to provide for study and research looking 
toward the improvement of forecasting and the general develop- 
ment of the science of meteorology. As the practical services in- 
creased, and as the science developed and required greater training 


and specialization of those in its service, it became evident that the 
demands of the work could be met more fully by the creation of 
an independent scientific organization, free from military regula- 
tions and devoting its entire attention to meteorology and climatol- 
ogy. Accordingly, effective July 1, 1891, the Congress established 
the Weather Bureau in the United States Department of Agricul- 
ture, and the entire official weather service was transferred to this 
Bureau, where it remained until June 30, 1940, when Congress trans- 
ferred it to the Department of Commerce. It has continually grown 
in scientific attainment and practical usefulness. 

Present Organization of the Weather Bureau 

The central office. The Weather Bureau organization consists of 
a central administrative and scientific office at Washington, D. C., 
and numerous offices and stations of various grades throughout the 
nation, including Alaska, Hawaii, and Puerto Rico. It is adminis- 
tered by a Chief of Bureau, selected for scientific and executive at- 
tainments and appointed by the President. The appointment is per- 
manent, with Civil Service status. All other commissioned employees 
are appointed after passing examinations given by the United States 
Civil Service Commission and serve under the regulations of 
the Civil Service acts. They are allocated to various professional, 
subprofessional, and clerical grades. Appointments are usually made 
to the lower grades, and the higher positions are filled by promotion. 

In the central office, under the Chief of Bureau, the technical 
services of the bureau are organized in several divisions, dealing 
with such services as ( 1 ) planning, organizing, and operating the 
network of Weather Bureau stations; (2) supervising communica- 
tion schedules and the preparation and distribution of maps, fore- 
casts, and warnings; (3) supervising the purchase and testing of 
meteorological instalments, and their installation, operation, and 
maintenance; (4) checking, summarizing, and publishing weather 
observations and climatic data and reporting effects of weather on 
crops; and (5) forecasting river stages and floods and maintaining 
a hydrologic program. Other divisions deal with scientific services 
such as conducting research, editing publications, and managing 
the largest, most nearly complete meteorological library in the 
United States, There is also necessarily an administrative organiza- 


tion concerned with such matters as personnel management, budgets 
and accounts, printing, and the procurement and distribution of nec- 
essary supplies. 

The field organization. The organization outside of Washington 
consists of four administrative regions, with a regional director in 
charge of each, over two hundred first-order stations, and a large 
number of lesser stations of many different classes. The regional 
offices serve as co-ordinating agencies between the central office and 
the field stations, and between the various types of service. They 
handle many administrative details in connection with supplies, 
equipment, inspection, and personnel. The first-order stations are 
local public offices of the Weather Bureau, manned by from one to 
several professional and subprofessional employees, and at these sta- 
tions complete meteorological records are kept. Most of them send 
daily telegraphic reports of observations and issue local forecasts. 
Many offices are at airports; in other cases, city offices are main- 

The numerous substations fill a great variety of special needs. 
Substations are divided into the following classes: second-order sta- 
tions, maintained primarily as weather observation stations, make 
daily telegraphic reports for the forecast and warning service; third- 
order stations telegraph daily weather observations at certain times 
for special purposes; river substations make and forward river-stage 
and precipitation observations; snowfall substations make snow- 
depth and sometimes snow-density measurements; display substa- 
tions display storm or hurricane signals or disseminate forecasts and 
warnings; crop substations make and telegraph observations for 
weather and crop bulletins or for frost forecasts; climatological sub- 
stations make observations for record or climatological purposes but 
do not telegraph them; airway substations make a record of weather 
conditions along the airways and transmit reports at stated times; 
co-operative substations are maintained primarily for climatological 
purposes and make daily observations of rainfall or of temperature 
and rainfall and send monthly reports by mail to section centers. 

Activities of the Weather Bureau 

The several activities of the Weather Bureau may all be classed 
under the head of the accumulation and distribution of weather in- 


formation. They have grown hi response to the needs and demands 
of the public and the ability of the Bureau to meet these needs, to 
a degree at least. The services provided have become an essential 
part of the daily life of the United States, and there is continual 
demand for increased and more detailed information. Meteorolog- 
ical services and meteorological research expanded greatly during 
the war and are being continued at higher than prewar levels. 

Forecast and warning service. The synoptic observations made 
at 6-hour intervals, and the upper-air observations made at selected 
stations are immediately transmitted in code over teletype circuits. 
They thus become available at Washington and at other forecast 
centers and first-order stations where forecasts are made. From these 
reports the weather maps are prepared according to a uniform pro- 
cedure and with uniform symbols, so that one familiar with the 
practice can readily interpret a weather map wherever it may have 
been prepared. The Washington office issues a large, lithographed 
daily weather map with supplementary charts and other data. Nu- 
merous other stations prepare maps according to their needs. Maps 
are also sent from Washington by a telegraphic process and printed 
in many newspapers. In connection with the maps, the Weather 
Bureau furnishes forecasts, weather summaries, and tabular matter. 
Weather maps of the north Atlantic Ocean are prepared in the New 
York office, and of the north Pacific in San Francisco. 

For forecast purposes, the country is divided into 14 forecast dis- 
tricts, each comprising a few or several states, corresponding in a 
general way with topographic and climatic regions. Each of these 
districts has a forecast center at which several forecasters are on 
duty, giving continuous 24-hour service. Forecasts are prepared 
every six hours, a separate forecast for each state and sometimes for 
portions of states. 

In recent years there has been increasing use of upper-air data 
and increasing skill in the interpretation of upper-air charts, as well 
as of frontal movements and their effects on the weather. Upper- 
air analysis has led to the development and regular publication of 
extended forecasts, as described in Chapter 12. 

At many first-order stations, the official in charge makes a daily 
local forecast for his vicinity, amplifying the state forecast, making 
it more specific, or occasionally differing from it. The significant 
features of the daily analysis, as it is made at forecast centers by ex- 


perienced, skilled forecasters after study of all relevant data and sup- 
plementary charts, are transmitted to first-order stations as a guide 
in the preparation of maps and the distribution of weather informa- 
tion. These are colloquially called canned analyses. Forecasts are 
given wide distribution by press associations, newspapers, radio 
broadcasting companies, and telephone companies. The officials in 
charge at Juneau, Honolulu, and San Juan issue forecasts for Alaska, 
Hawaii, and Puerto Rico, respectively. 

Special forecasts or warnings are issued either in connection with 
the regular daily forecasts or at other times when injurious or haz- 
ardous weather conditions are expected. Such are the storm fore- 
casts and warnings made by the Severe Local Storm Center when 
tornadoes, other damaging winds, or hailstorms are imminent. An- 
other specially organized unit provides a hurricane warning service 
for the Caribbean Sea and along our Gulf and south Atlantic coasts. 
Special forecasts of heavy snow, high winds, low temperatures, and 
other hazardous conditions are provided for the general public when 

Climatological service. , The first-order stations are placed pri- 
marily with a view to serving as a network for forecasting purposes 
and as centers for the dissemination of weather news to the public. 
They accumulate much valuable climatological data but are not suf- 
ficiently numerous to cover the local climatic variations found in 
each state. To extend climatic studies to smaller areas and to estab- 
lish the main features of the local climates of the country, the clima- 
tological service of the Weather Bureau has been organized. Clima- 
tological data are obtained by a large number of co-operative sub- 
stations, well distributed, so that most of the counties of the United 
States are represented. There are about 5,600 such stations in the 
country, or about 1 to every 550 square miles of area. The stations 
are often only from 20 to 30 miles distant from each other. 

Co-operative stations are not equipped with recording instru- 
ments, but each has an 8-inch rain gage, and the majority also have 
an instrument shelter, inclosing maximum and minimum thermom- 
eters. They are thus equipped to obtain the climatic elements of pri- 
mary importance, that is, the daily maximum and minimum tem- 
peratures and the daily rainfall. The co-operative observer, whose 
work is voluntary and gratuitous, reads the thermometers and rain 
gage once each day. He records the readings and certain other ob- 


servations on a standard form which he mails to a processing center 
at the end of the month. The material thus collected is published 
in detail in a monthly bulletin, called Climatological Data. The final 
repository for American weather records, plus electronic computers 
for statistical computations, is maintained at the National Weather 
Records Center in Asheville, North Carolina. 

Agricultural and horticultural service. The special agricultural 
service of the Weather Bureau consists largely in the publication of 
bulletins containing weather data and reports on the effect of the 
weather on the condition and progress of crops. The central office 
at Washington publishes a Weekly Weather and Crop Bulletin 
throughout the year, with reports and summaries for all parts of 
the country. During the winter, this bulletin also contains reports 
of the depth of snow on the ground and the thickness of ice in rivers 
and harbors. 

In addition, daily bulletins contain reports from the crop substa- 
tions and are named according to the principal crops grown in the 
areas represented. They are bulletins for the corn and wheat region, 
and for the cotton, fruit, sugar, rice, and cranberry regions. The 
horticultural service gives warnings of frosts likely to be injurious 
to fruits in those regions where fruit-growing is a major industry 
and where protective measures are practicable. 

Marine meteorological service. About 2,000 ocean-going ships 
make weather observations at Greenwich mean noon each day, in 
whatever part of the world they may be. Large numbers of records, 
coming from the masters of vessels of every maritime nation, are 
received at the central office of the Weather Bureau and are there 
studied and summarized. These ships' records are the basis of our 
knowledge of the climates of the oceans. The summarized and 
charted data are furnished by the Weather Bureau to the hydro- 
graphic office of the Navy Department and are published by that 
office on its monthly pilot charts, used by all navigators. Wheii 
within certain areas, many ships transmit daily observations by 
radio. These reports make possible the daily weather maps of the 
north Atlantic and north Pacific oceans and are also valuable in 
locating and forecasting the course of West Indies hurricanes. 

Coast Guard patrol vessels make regular weather observations, 
and in recent years the use of "weather ships" in the north Atlantic 
and north Pacific oceans has become a common practice. These 


ships are assigned duty stations along the major ocean airline routes 
and remain on station for several days before being relieved by an- 
other vessel. They make regular surface and upper-air observations, 
but they may occasionally be called from their station to serve as a 
rescue vessel in case of a disaster. 

Aviation weather service. The Aviation weather service operates 
24 hours every day to supply weather information and forecasts for 
air operations. About 260 domestic airport stations are connected 
by teletype circuits. They collect and exchange information about 
the weather every hour. Individual flight weather briefing or route 
forecast service is available at all airport stations. The aviation 
weather service is operated in co-operation with the Civil Aero- 
nautics Administration. 

Fire-weather warning service. The control of forest fires is a dif- 
ficult problem involving heavy expenditure of time and money and, 
often, the services of large forces of men. The efficiency of fire con- 
trol is largely dependent upon forestry officials' being prepared for 
emergencies as they arise. The forest fire-weather warning service 
provides forecasts of humidity, wind, and thunderstorm conditions 
in the forested regions of the country to assist other agencies in 
combating the fire menace. 

Hydrologic services. The principal rivers and tributaries of the 
United States are observed daily at about 1,000 stations, and daily 
river-stage forecasts from one to three days in advance are made 
for the principal rivers. These forecasts are especially valuable for 
navigation, water utilization, and flood control. 

In co-operation with the United States Corps of Engineers and 
the Bureau of Reclamation, the Weather Bureau collects and pub- 
lishes data from a network of some 3,000 special precipitation- 
measuring stations. Recording rain gages are used in this effort to 
determine storm intensity over a given watershed. 

A network of snowfall stations is maintained in the mountains of 
the western states during the winter and spring months. These sta- 
tions supply the basis for forecasts, for irrigation and other water 
interests, of the amount of runoff from melting snow during the sum- 
mer and fall months. 

Research and publications. The Weather Bureau encourages 
special studies devoted to improving the accuracy and efficiency 
of its many forecasts. Recent research activities have centered pri- 


marily on (1) improved instrumentation, (2) long-range forecast- 
ing, (3) the jet stream, (4) circulation patterns in the high atmos- 
phere, (5) tornado studies, and (6) the processes that attend con- 
densation of moisture, including investigations of the merits of 
artificial rain stimulation. 1 

Research findings are made available in the Weather Bureau Re- 
*earch Papers and the Monthly Weather Review. A number of other 
publications are made available, at regular and irregular intervals, 
to those interested in particular aspects of the Weather Bureau's 
work. Regular publications include the Daily Weather Map, Weekly 
Weather and Crop Bulletin, and Climatological Data, in addition 
to the Monthly Weather Review. Irregular publications include the 
above-mentioned technical papers, various instructional circulars, 
Frost Charts, and other miscellaneous items. 

Military and Private Weather Services 

Because the military forces often operate beyond the regions 
served by Weather Bureau forecasts, and because their needs often 
demand particular types of forecasts, the armed forces have created 
their own weather services. The growing importance of aircraft in 
modern warfare and the importance of naval operations in military 
activities of the United States have increased the necessity for mili- 
tary weather units. The Air Force and the Navy maintain weather 
services for their own activities, and it is reported that the United 
States Army Signal Corps is planning to develop a weather service 

Air Force and Navy weather services. The United States Air 
Force Air Weather Service was organized to participate in weather 
observations, prepare forecasts, and perform research for the benefit 
of the armed forces, particularly for the aid of the aviation activities 
of the Air Force. Units of the Air Weather Service are assigned to 
stations wherever meteorological services are required, and may be 
transferred from place to place as the need arises. When operating 
in the United States, the Air Weather Service co-operates with the 
Weather Bureau and in general uses their facilities if possible. The 

1 Sinclair Weeks, "Weather Bureau," United States Government Organization 
Manual 1953-1954. Washington, D. C.: Government Printing Office, July 1, 1953, 
pp. 285-288. 


United States Air Force took the lead in establishing facsimile cir- 
cuits for weather transmission after the medium became practical 
during World War II (Fig. 145). It also has sponsored much valu- 
able research in recent years, notable among which is the work on 
tornado forecasting discussed in Chapter 11. 

big. 145. facsimile, This instrument reproduces maps ana cnarts trom a master 
copy drawn in Washington, D. C. It requires a line connection similar to telegraph. 
A modification of this machine can reproduce maps from radio transmissions. Courtesy, 
Times Facsimile Corp., New York, New York. 

The Aerology Section of the United States Navy is primarily con- 
cerned with supplying basic forecasts and wamings for its aircraft 
and for its ships at sea. A weather central at San Francisco collects 
and disseminates weather information to the North Pacific area by 
radio. A similar weather central at Norfolk serves the North Atlan- 
tic area. Smaller weather units or single individuals are assigned to 
ships or activities, according to their needs for weather service. The 
Navy, like the Air Force, has been busily engaged in research ac- 
tivities since 1946. Notable among them have been forecasting tech- 
niques for hurricanes and typhoons, the jet stream, and long-range 
forecasting. The Navy and the Air Weather Service have co-operated 
in hurricane and typhoon reconnaissance over the subtropical oceans 


to such an extent that property damage and loss of life have been 
materially reduced. 

Private weather services. A complete new type of weather serv- 
ice has developed in the United States since 1945. This was pri- 
marily the result of about 6,000 young men being trained in mete- 
orology by the armed forces during World War II. 2 Before the war, 
there were only a very few persons engaged in private meteorology, 
most of them employed by airlines to supplement the services ren- 
dered by the Weather Bureau. 

Young men, usually with college degrees in mathematics or phys- 
ics, were chosen for the military training program in meteorology. 
They served as weather forecasters during the war and subsequently 
were released to inactive duty. Many of these men set themselves 
up as meteorological consultants and found a market for their serv- 
ices. At the present time, there is no separate, formal registration 
required of private professional meteorologists", so there is no accu- 
rate way to estimate the total number in private practice. There 
were only a few hundred professional meteorologists in the United 
States in 1940, but it is probable that there are that many currently 
engaged in private practice. 

Prospect for the Weather Services 

Co-operation has been effectively practiced among the various 
services since 1940. The Weather Bureau has prided itself on co- 
operating with everyone, including many other branches of the 
federal government, the armed forces, and private meteorologists. 

Numerous research projects and operational techniques have been 
jointly subsidized by the Weather Bureau, the Air Force, and the 
Navy. Probably the best example of this is the establishment of 
WBAN (Weather Bureau, Air Force, Navy), a master weather- 
analysis group in Washington, D. C., WBAN distributes analyzed 
weather charts of both surface and upper-air data by means of fac- 
simile and by coded teletype transmissions. This tends to co-ordinate 
field analyses and saves manpower. It has possibilities of greatly 
improving the weather services. 

The private meteorologist, as a professional weather forecaster, 

2 United States Employment Service, Meteorology as a Profession. Washington, 
D. C.: Department of Labor, 1946, p. 11. 


has created difficulties of administration among the weather serv- 
ices. These difficulties will disappear when the areas in which 
private meteorologists are to function have been clearly defined. Re- 
quests for Weather Bureau services have increased in such numbers 
and in such varieties of information desired as to impair the normal 
functions of the Bureau. Private meteorologists stand ready to fur- 
nish these services for special interests and in all probability will 
contribute toward better weather services for all. 

New ways are being recognized or developed in which weather 
services become more valuable to everyone. Television meteorolo- 
gists are effectively developing a greater interest in weather and in 
the importance of reliable weather information. In this modern age, 
when industrial efficiency is requiring more accurate weather fore- 
casts and when meteorology is on the threshold of becoming a more 
exact science, the future of the meteorologist seems rather bright. 



General Elementary Treatises 

Albright, J. G., Physical Meteorology. Englewood Cliffs, N. J.: Prentice- 
Hall, Inc., 1939. 

Brands, G. J., Meteorology, A Practical Course in Weather. New York: 
McGraw-Hill Book Company, Inc., 1944. 

Donn, William L., Meteorology with Marine Applications. New York: 
McGraw-Hill Book Company, Inc., 1946. 

Haynes, B. C., Meteorology for Pilots. Washington: Government Print- 
ing Office, 1943. 

Kendrew, W. G., Weather: An Introductory Meteorology. New York: Ox- 
ford University Press, 1943. 

Mudge, Robert W., Meteorology for Pilots. New York: McGraw-Hill 
Book Company, Inc., 1945. 

Neuberger, Hans H., and F. Briscoe Stephens, Weather and Man. Engle- 
wood Cliffs, N. J.: Prentice-Hall, Inc., 1948. 

Petterssen, Sverre, Introduction to Meteorology. New York: McGraw- 
Hill Book Company, Inc., 1941. 

Tannehill, Ivan Ray, Weather Around the World. Princeton: Princeton 
University Press, 1943. 

Taylor, George F., Elementary Meteorology. Englewood Cliffs, N. J.: 
Prentice-Hall, Inc., 1954. 

Advanced Treatises involving Considerable Mathematics 

and Physics 

Berry, F. A., E. Bollay, and N. R. Beers, Handbook of Meteorology. New 

York: McGraw-Hill Book Company, Inc., 1945. 
Byers, Horace R., General Meteorology. New York: McGraw-Hill Book 

Company, Inc., 1944. 



Garbell, Maurice A., Tropical and Equatorial Meteorology. New York: 

Pitman Publishing Company, 1947. 
Haurwitz, Bernhard, Dynamic Meteorology. New York: McGraw-Hill 

Book Company, Inc., 1941. 

Hewson, E. W., and R. W. Longley, Meteorology, Theoretical and Ap- 
plied. New York: John Wiley and Sons, Inc., 1944. 
Holmboe, J., G. E. Forsythe, and W. Gustin, Dynamic Meteorology. New 

York: John Wiley and Sons, Inc., 1945. 
Humphreys, W. J., Physics of the Air, 3rd ed. New York: McGraw-Hill 

Book Company, Inc., 1940. 
Johnson, John G, Phijsical Meteorology. New York: John Wiley and Sons, 

Inc., 1954. 
Malone, Thomas F., Compendium of Meteorology. Boston: American 

Meteorological Society, 1951. 
Petterssen, Sverre, Weather Analysis and Forecasting. New York: 

McGraw-Hill Book Company, Inc., 1940. 
Riehl, Herbert, Tropical Meteorology, New York: McGraw-Hill Book 

Company, Inc., 1954. 
Starr, Victor P., Basic Principles of Weather Forecasting. New York: 

Harper and Brothers, 1942. 

Sutton, O. G., Micrometeorology. New York: McGraw-Hill Book Com- 
pany, Inc., 1953. 
Taylor, G. F., Aeronautical Meteorology, 3rd ed. New York: Pitman 

Publishing Corporation, 1941. 
Willett, Hurd C., Descriptive Meteorology. New York: Academic Press, 

Inc., 1944. 

Subdivisions of Meteorology and Related Topics 

Ashley, Frances D., Weather Horizons. Boston: American Meteorological 

Society, 1947. 
Bentley, W. A., and W. J. Humphreys, Snow Crystals (contains over 

2,000 microphotographs). New York: McGraw-Hill Book Company, 

Inc., 1931. 
Berriman, A. E., Historical Meteorology. New York: E. P. Button and 

Company, Inc., 1953. 
Brier, Glenn W., A Study of Quantitative Precipitation Forecasting in 

the TV A Basin, Weather Bureau Research Paper No. 26. Washington: 

U. S. Department of Commerce, 1946. 
Byers, H. R., and R. R. Braham, Jr., The Thunderstorm. Washington: 

U. S. Department of Commerce, 1949. 
Byers, Horace R., Thunderstorm Electricity. Chicago: The University of 

Chicago Press, 1953. 
Caudle, Frederick L., Workbook in Elementary Meteorology. New York: 

McGraw-Hill Book Company, Inc., 1945. 
Clayton, H. H., World Weather Records, Vols. 79 (1927), 90 (1934), and 


105 (1947) of Smithsonian Miscellaneous Collections. Washington: 
Smithsonian Institution. 

Coons, R. D., E. L. Jones, and R. Gunn, Artificial Production of Precipita- 
tion, Weather Bureau Research Paper No. 33. Washington: U. S. De- 
partment of Commerce, 1949. 

Elliot, R. D., Extended Weather Forecasting by Weather Type Methods. 
Washington: U. S. Weather Bureau, 1944. 

Flora, Snowden D., Tornadoes of the United States. Norman, Oklahoma: 
University of Oklahoma Press, 1953. 

Goody, R. M., The Physics of the Stratosphere. Cambridge: Harvard 
University Press, 1954. 

Haynes, B. C., Techniques of Observing the Weather. New York: John 
Wiley and Sons, Inc., 1947. 

Humphreys, W. J., Fogs, Clouds and Aviation. Baltimore: Williams and 
Wilkins Company, 1943. 

International Meteorological Committee, International Atlas of Clouds 
and of States of the Sky. Paris: Office National M6t6orologique, 1932. 

Johnson, O. J., Correlation of Cycles in Weather, Solar Activity, Geomag- 
netic Values, and Planetary Configurations. San Francisco: Phillips 
and Van Orden Company, 1946. 

Marvin, C. F., Psychrometric Tables, Weather Bureau Bulletin No. 235. 
Washington: U. S. Department of Commerce, 1941. 

Middleton, W. E. K., Vision Through the Atmosphere. Toronto: Uni- 
versity of Toronto Press, 1952. 

Middleton, W. E. K., and A. F. Spilhaus, Meteorological Instruments, 3rd 
ed. Toronto: University of Toronto Press, 1953. 

Muller, Siemori W., Permafrost; or Permanently Frozen Ground and Re- 
lated Engineering Problems. Ann Arbor, Michigan: J. W. Edward, 
Inc., 1947. 

Namias, Jerome, Methods of Extended Forecasting. Washington: U. S. 
Weather Bureau, 1943. 

Newell, Homer E., Jr., High Altitude Rocket Research. New York: 
Academic Press, Inc., 1953. 

Perrie, D. W., Cloud Physics. Toronto: University of Toronto Press, 1951. 

Pulk, E. S., and E. A. Murphy, Workbook for Weather Forecasting. Engle- 
wood Cliffs, N. J.: Prentice-Hall, Inc., 1950. 

Schonland, B. F. J., Atmospheric Electricity. New York: John Wiley and 
Sons, Inc., 1953. 

Spilhaus, A. F., and J. E. Miller, Workbook in Meteorology. New York: 
McGraw-Hill Book Company, Inc., 1942. 

Sverdrup, H. U., Oceanography for Meteorologists. Englewood Cliffs, 
N. J.: Prentice-Hall, Inc., 1942. 

Tannehill, Ivan Ray, Hurricanes: Their Nature and History, 5th ed. 
Princeton: Princeton University Press, 1944. 

Thiessen, Alfred H., Weather Glossary. Washington: Weather Bureau, 
U. S. Department of Commerce, 1946. 


Thornthwaite, C. W,, and B. Holzman, Measurement of Evaporation from 

Land and Water Surfaces, Tehnical Bulletin No. 817. Washington: U. 

S. Department of Agriculture, 1942. 
Vaeth, J. Gordon, 200 Miles Up: The Conquest of the Upper Air. New 

York: The Ronald Press Company, 1951. 
Weightman, R. H., Average Monthly Tracks by Types of Lows in the 

United States. Washington: U. S. Weather Bureau, 1945. 

Climatology and Related Topics 

Aronin, Jeffrey Ellis, Climate and Architecture. New York: Reinhold Pub- 
lishing Corporation, 1953. 

Blair, T. A., Climatology, General and Regional. Englewood Cliffs, N. J.: 
Prentice-Hall, Inc., 1942. 

Brooks, C. E. P,, Climate in Everyday Life. New York: Philosophical 
Library, 1951. 

, Climate Through the Ages. New York: McGraw-Hill Book Com- 
pany, Inc., 1949. 

Brooks, C. F., A. J. Conner, and others, Climatic Maps of North America. 
Cambridge: Harvard University Press, 1941. 

Conrad, V., and L. W. Pollak, Methods in Climatology, 2nd ed. Cam- 
bridge: Harvard University Press, 1950. 

Geiger, Rudolf, The Climate Near the Ground, translated by M. N. Stew- 
art and others. Cambridge: Harvard University Press, 1950. 

Hadlow, Leonard, Climate, Vegetation and Man. New York: Philosophi- 
cal Library, 1953. 

Hare, F. K., The Restless Atmosphere. London: Hutchinson's University 
Library, 1953. 

Haurwitz, B., and J. M. Austin, Climatology. New York: McGraw-Hill 
Book Company, Inc., 1944. 

Jacobs, Woodrow C., Wartime Developments in Applied Climatology, 
Meteorological Monographs, Vol. I, No. 1. Boston: American Meteor- 
ological Society, 1947. 

Kendrew, W. G,, Climates of the Continents, 4th ed. New York: Oxford 
University Press, 1953. 

Landsberg, Helmut, Physical Climatology, 2nd ed. State College, Pa.: 
Pennsylvania State College, 1942. 

Markham, S, F,, Climate and the Energy of Nations. New York: Oxford 
University Press, 1947. 

Mills, C. A., Climate Makes the Man. New York: Harper and Brothers, 

Shapley, Harlow, Climatic Change. Cambridge: Harvard University 
Press, 1953. 

Trewartha, Glenn T., An Introduction to Climate, 3rd ed. New York: 
McGraw-Hill Book Company, Inc., 1954. 


nited States Department of Agriculture, Climate and Man, 1941 year- 
book. Washington: Government Printing Office, 1941. 


mateur Weatherman's Almanac, published annually by Weatherwise, 
Franklin Institute, Philadelphia, Pennsylvania. 

verage Monthly Weather R6sum6 and Outlook, published semimonthly 
at Washington by the U. S. Weather Bureau. 

ulletin of the American Meteorological Society, published monthly ex- 
cept July and August by the American Meteorological Society, Boston, 

limatological Data (by states), monthly bulletins and annual summary 
obtainable from the U. S. Weather Bureau Section Directors. 
>aily Weather Map, published at Washington by the U. S. Weather 

ntrnal of Meteorology, contains original and highly scientific papers; 
published bimonthly by the American Meteorological Society, Boston, 

teteorological Abstracts and Bibliography, contains abstracts and bib- 
liography of meteorological research all over the world; published 
monthly by the American Meteorological Society, Boston, Massachu- 

lonthly Weather Review, contains technical contributions in synoptic 
and applied meteorology and a monthly weather summary; published 
by the U. S. Weather Bureau, Washington, D. C. 
Quarterly Journal of the Royal Meteorological Society, contains the re- 
sults of original research; published by the Royal Meteorological So- 
ciety, London, England. 

feather, monthly magazine published by the Royal Meteorological So- 
ciety, London, England. 

featherwise, a bimonthly magazine by Amateur Weathermen of Amer- 
ica, Franklin Institute, Philadelphia, Pennsylvania. 
^eekly Crop Bulletin, published at Washington by the U. S. Weather 

Miscellaneous Publications of the U. S. Weather Bureau *: 

Circular A. Instructions for obtaining and tabulating records from 
jcording instruments. 

Circular B. Instructions for Co-operative Observers. 

Circular C. (Combined with Circular B.) 

Circular D. Instructions for the installation and maintenance of wind- 
leasuring and -recording apparatus. 

* Obtainable from the Superintendent of Documents, Government Printing Office, 
Washington, D. C. 


Circular E. Measurement of precipitation. 

Circular F. Barometers and the measurement of atmospheric pres- 

Circular G. Care and management of electrical sunshine recorders. 

Circular I. Instructions for erecting and using Weather Bureau neph- 

Circular L. Instructions for the installation and operation of class A 
evaporation stations. 

Circular M. Instructions to marine meteorological observers. 

Circular N. Instructions for airway meteorological service. 

Circular O. Instructions for making pilot-balloon observations. 

Circular P. Instructions for making aerological observations. 

Circular Q. Pyrheliometers and pyrheliometric observations. 

Circular R. Preparation and use of weather maps at sea. 

Circular S. Codes for cloud forms and states of the sky. 

Circular T. Ocean station instructions for meteorological personnel 
( supplementary ) . 

Cimatic Charts for the United States. 12 charts, 10 X 16 inches. 

Cloud Forms. Descriptive pamphlet and 32 halftone plates. 

Frost Charts for the United States. 5 charts, 10 X 16 inches. 

Weather Code. Includes Synoptic Code, Radiosonde and Rawinsonde 
Code, Upper Wind Code, U. S. Weather Analysis Code, and sheet con- 
taining Station Model and Explanation of Weather Code Figures and 



Abridged from Smithsonian Meteorological Tables, Fifth Edition, 

Equivalent Values 

1 foot 0.3048 meter. 

1 meter = 39.37 inches = 3.2808 feet. 

1 mile = 1.6093 kilometers. 

1 kilometer = 3280.8 feet = 0.62137 mile. 

1 inch, mercury = 25.4 millimeters = 33.86395 millibars. 
1 millimeter, mercury = 0.03937 inch = 1.3332 millibars. 
1 millibar = 0.02953 inch 0.75006 millimeter. 

1 mile per hour = 1.467 feet per second 
1 mile per hour = 0.447 meter per second 
1 mile per hour = 1.610 kilometers per hour 
1 mile per hour = 0.868 knot 

1 meter per second 2.237 miles per hour 

1 meter per second = 3.600 kilometers per hour 

1 meter per second = 1.940 knots 

1 knot = 1.152 miles per hour 

1 knot = 1.854 kilometers per hour 

1 knot = 0.515 meter per second 


F=% C + 32 =%(A-273)+32. 
A=%(F-32) + 273 m C + 273. 


































































































































































































































































































































































30 1016 1019 1023 1026 1029 1033 1036 1040 1043 1046 

31 1050 1053 1057 1060 1063 1067 1070 1073 1077 1080 





Statute Nautical 

Miles Miles Feet Kilometers 

1 0.9 5,280 2 

2 1.7 10,560 3 

3 2.6 15,840 5 

4 3.5 21,120 6 

5 4.3 26,400 8 

6 5.2 31,680 10 

7 6.1 36,960 11 

8 6.9 42,240 13 

9 7.8 47,520 14 

10 8.7 52,800 16 

11 9.6 58,080 18 

12 10.4 63,360 19 

13 11.3 68,640 21 

14 12.1 73,920 23 

15 13.0 79,200 24 

16 13.9 84,480 26 

17 14.8 89,760 27 

18 15.6 95,040 29 

19 16.5 100,320 31 

20 17.4 105,600 32 

21 18.3 110,880 34 

22 19.1 116,160 35 

23 20.0 121,440 37 

24 20.9 126,720 39 

25 21.7 132,000 40 

26 22.6 137,280 42 

27 23.5 142,560 43 

28 24.3 147,840 45 

29 25.2 153,120 47 

30 26.1 158,400 48 

31 27.0 163,680 50 

32 27.8 168,960 51 

33 28.7 174,240 53 

34 29.6 179,520 55 

35 30.4 184,800 56 

36 31.3 190,080 58 

37 32.2 195,360 60 

38 33.0 200,640 61 

39 33.9 205,920 63 

40 34.7 211,200 64 

41 35.6 216,480 66 

42 36.4 221,760 68 

43 37.2 227,040 69 

44 38.1 232,320 71 

45 39.0 237,600 72 

46 39.9 242,880 74 

47 40.8 248,160 76 

48 41.6 253,440 77 

49 42.5 258,720 79 

50 43.4 264,000 80 

+ I I I I I I I 

I I I I I I I t I I I 

I I I I I I I I f I I 

i i i i i i i i i 

I I i I I t I I I I 

I I I I I 1 I I I I 

I I I 1 I I I I I 

O H 


S319V1 QNV 






t> O 
C H 

H 3 





1 1 1 1 I 

I I 77 ^ 


1 1 1 1 


1 1 1 1 

TTT I ' 


I I I I ' 

1 1 1 1 1 1 + 








Q. 3 

Q ~ 


^V >H 

U, g 





a a 

< 2 



>- S 

I y 







U51O Of- OOO OCO C00 

co OO^H <N . ?'! 


"tO O'f H"tf IO Nt 
10 "? ^'J H^ CO'I 0> 
WN ^<O UDH U5CO "* 


id 1 ^ od^. H"? co'^l Q*. 
ON JOH coo coco coco 




SON -f N ^N coo* COH 


coco co ^ 10 o oo coo 


OS^ r>jO .^O _JlO 

CON ^d So cod 
^2S ""19 *S ^.29 

/rvjW to ? **Q tsi^ 


N <(N . t^^ H H . 0>^ t^ H <U ? N 1 ^ CO <C 1 O"*. CO 1 ^ O'^^UD^OS^ 
OON OH t>O OCO COCO 00<0 C^CO I>CO t-H |>00 t>N 1>N CO O CON 

OON 00^ <fif t*0> COO 
N <r 1o>^N^H^O^ 



N? ^ <CO , o >Tl ! o' 1 "! co ^ 


^H HN TfOO Tftf COlO 
^J O ,,] IO Q H >X CO ^2 H 

S iCq . rjJ9 N >C % IO "^ O . 

H C0 ^^ U3 10 . 

0>05 NO 0>H COOO COCO 

od^od^o^co^d 1 ^ 


OO 10 ^r coo NH coco 

oco toco' coco ION co-^ 



HN ^ ^f ^rf 000 
N ? 00 H . UJ H . nl 



S <C 1 CO 00 . H t<N . C0? 


NCO COTj< ^N H^ 

^N IOH TfH "tN 

<<*<N COOO HO CO ^| 0> 00 

o' s , co 10 . 05"? o>. ui 1 ^ 


^f o coo Tf o OH 
co^. N9 M 1 *: od? 


| OOO rj< O Tj<00 ^Tf OO OCO ON 

NCO cqio oqco O,H coco ON HO coco oqo HO ^o OH NO wco 


i ^j rf| ^| tiajg 




COt *^ lO CO O CO ^* CM O CM ^ CO IOCO IO rH COCM CO ^ 000) **rH COt* C*D 00 rH 

'o* *i> CM*CM* CMOO CMIO t-co ^'id *t- cdrn c&t- ot- CMCM* 0*1-4 dt oJc- cMcd 

<* LOCO <*rH CD* COrH IQ * C* IO ^ CM $10 3 * Cp IO U) <* IOCO IOCM CO SjcO 

ScO* COrH CMO *'** ScM* COCO* CO rH rH O* * ^ "* * S ^" CO CO CO rH CM O S rH C? CM* 

CMCO t CO COO rH 0) Oft O t^ rH COrH Tf t 00 CO OO IO ^* CMCO CO t IO t* C*" O CM O) 

COrH TfrH COO O ^' CD rH * CO t- CM COrH IO CO ^ CO COCO ^ CM* ^ rH CT)rH IO O ^ CM 

rHCO t^(N IOCO COrH COCO CO IO O "^ 9$ 9 CO 9^ ^ COCO tOCO COt- CO t- t^CM 


OQ-CO Q^lO ^'05 ^'rH Qj'rH o 't^ Q *J rH 1 CO 9 rH^ QJ 9 CO* ? CM 9 CO ! CM* ^ CO *3 


coid i>*coot>co*t>o'i>co' coco coco co Tf't-cdcoid S-^t-cM't-co'SdSco 

oo cocn icio -^co 
i 00 . co*. 0*1 c^^. 
-io t--tf coin t*co 

co C^CM oco CCM coo coo coco 
' s . ^o ^jcq ^CM ^oq ^q 6 5J 
o i>co coco CD-** rco t*"* cow 

Or- coc 
^-co ^i 
c-oo t- 

B t ** TH 

o Scd 

t- 1 * CO 1 * U5CM t* COO COCO C-CO IOCO l>co' COCO t>* COCO CO* COCO t-O COCO 

S'tt *' r "lo$9cM lH .oS. CO. M. cd^ld^oi^lod^ ai^aJ^lrH 1 ^ t>. rH^ 


COrH t-00 *O> C>CM IO 00 IO O> CM t* COCM 00 Cft CM rH COCM t> * O 00 O fr- C^ 00 COCO 

COCM * CM COrH 10* IOCM** t-CM'cMrHl5ld*lo'cO* COCoSrHCOrH'SoScO 





JfS nS ; ij-jjf-g ;j 


..... ,j ..... . 

-^ : j : ; ; y 






^f W CD CO ^ CO O W t- ' CO CO W CO ^ 

LOO> HCO HW P coco W'* COH qco HO oqw woo qq wco OH <<*io 


^^ IO* CO-^f -^00 U3CO IOH COH LOW ^ H COCO LOCO * H LO f ^ 

t^> H 1 ^ CO 2 ? 0>> '<". 
WOO ^<D ^00 WW COW 

0) CO O t CO LO t "^ LO ^J 

rH d"1 CO ^ 

8-^LO O- 
W r 1 H >C ? 


H 6 , 



qco ooo OOQ 
cp^. cd""! H*I 

~ co r 

co -^cb Low cow r?w 


oi 1 "! ^ 1H . w^ H^ 06^ 


HLO >co COLO ojop qg 


i'. S d*1 S* S^. 

irH COO COH LOH ^00 


coco to we- ^o 
i 9 oo 9 lo^ co 

rt<^ 00|> ^t"^ COW 
rH* 00 'I C^ 00 . 00^. 
IOW ""* rH tOW "^O 

oow oc 
cd9 w 6 

t-H ^ 


S'-J jn'Oi ^ oi'"! d"^ 

^.00 Hg C3i0 05W WOS 

Hl^ WO iHjdJ WrH 
00^ O9 00? O>^. 



<*;* t-^J OH WOO 10 C> 

co 1 -! cd9 i>*. o>q t^ 5 * 


IOU5 ^211 r4*2 ^00 iHW 
TfOO 00 9 OJ. r^^. W <t0 . 

t*d coo lao cow cow 

C0^00> 10. 0><". 

^o ooo 
06^ LO^ 

t-H LOC- 



id 1 ^ t^9 od. oi^ co 00 , co'9 ai 50 . 
t-o coo LOO cow cow coo cod 


co^ o>9 ^i> -w 




WOO ^W t-O> -4Q O 

S' w . w >l 3 id" co9 id 
CO 5OrH ^"^ COCO I>C 

COO OiOi WrH iH^< O 

co"*. o >r 1 co. i>* t^ 


'tf O 0>UJ U500 0000 COW 

c-i 00 . co'9 06*"! ^'. 06^ 
coo ci5o So Sr-t Sw 

^w oo 10 ooo t^co oco 
t:a> o*>. co oq ^w ^o 



0> W <C 1 W <rH . W 1 ^ 



rf0 tO^ "> CO 00 . 


coo wco 


GO co o co 

^. 00*1 
t*H "tfLO 

co jo w^ OH t^o -^o 
o>9 od^ id <. cd^ c^^. 

^W SO lOrH -rH COW 

4*00 "*CO COt- C-^< 
OS ? 00^. CO >(N . CO 00 . 

OOCO 0>H (Nt> 00<O iHOO 

,H <o'>. d 00 , TH^: TjJ" 

OOCO "tfCO U500 COW "tCO 

000 iHCg WW COO) OOCO 

co'9 w^ co^ TjJ? ^35 

Wrf ^*fl ^tCO WW COW 

ww cow 

H*9 cd 1 ^. TjJ H . oi 00 . i-**. 


OOH ^.22 w LO HLO q^* 


WrH COCO ><* t> CO 10 

o>? ^^ o>"^ oo 50 . co^ 

WrH IOW <*<* WO rHrH 


^9 oi^. w^. d 35 . 

rfCO OOH "tCO ^O 

HO> COt- COt- LOH 
H. H 1 ^ 10^ C0"9 

IOTT LOCO "fLO coco 
oi9 t^^. co od 
co^ ww coco WH 

-*co LOH 
t-co COLO 

ooco ot> 
d^ d^ 

t-Tf COLO 

d 1 ^ cd9 

tco wco 



1 ~3oS 


j CMO 

w^ oq CN WTJ< 

lOTH^d O> O5 
^ CO CO CM t>CM 

OTJJ oc* oiH 

<N O iH O> CO 
UJ t- W WCM 

CM oq coi> r>co 

CO t~ 00 W ^ 

CMCO COO ^.^ ^^ ^ T "j ^ "* 

tH t-W CM CM rH CM 

Trt>aqooqo ( co > o^co i oo > qioooa>t>rHioio cooqoqoocococoo 

S'O* WfH O5CO* I>CC) t><N C^'r-J O H <N O ^ CO CO* O* t*' CM* IO 2 C^' CM " ^ 

CO CM CO CO CM t>CN CO t> lO CO <O O C- ift 

3i t> 00 CO O> CO 00 t CO O r-J ^ C^l C0 f CO f CO ^ O Ctt 00 l> OJ ^JJ < 

U5CM ^Oi ^JCO t>W 00 CO HvH HO W t> CN ^H^.^rH rH O CO IO CO p OJ O CO 

S'lH odd id co coco OCM WCM* aicM wa> oco'tH^' t>co HO ooco JOCM' 
1C if) W CO Ub t>rH CO CO CO W CO t> LO 






S'cMoJTHcd-^oocoaJcococo oco'ujcxirHi 
CO CO W CO CO 00 rH CO 00 

CRN U5C^ CMi-J t^Oi CO O CO t O 


r- t> rH CN CO l> 
CO t> 


^9 oo 
coo t>i 

O>CM <NCO cooq CMCO o>q qg I>CN wo 


toco o>q iqco woo 
'' ' 


cMt>oqcs|qcocqcM t 


qco qa> 



TH^ CMCO cjjoq wa woi coco cot- OCM ^>Oi-jgt^cq 
dcoTfcdcMCNcod O>COHO'OCM^O coco j-fS ^2 

rHCNCMi f-COl>W t> l> O t* H 



iqo COO 


O f ^ O"tf 00 Cq 00 Oi 0t> 

83 S3 5SS 88 $8 

J 00 (O OO COH 



r- ot 


t^ 16 D 

0>0 C0t> OOO COO ^ 00 
COO OOiH C^rH l>C<i U3r4 

t> to 10 <x> 10 

wi> TH^J osco ooo t i-jo t>"<f oqiH wco 

O W O) ^T O> c4 "* iH GO Tj* OJrH iHtH QC4 

10 T* ^ S 25 CO W U5 



0>CO* t^ iH OrH 

10 10 co 

N f-i O> 

co 10 

f*H >CO OO rH*C<i "tCO COrH r4 O? O C^' I> CO* M* <N U5O' WO* 

tO "^ OOt>?OI>CD <DOCOt^t* 


' ' 


OCO OC4 t>O TH(N ^'oi rHrH O'rH r-i<N lOCO* COrH OrH iHrH C0 1-4 O> r4 <O < 

t> 10 co co to co 10 o 10 10 10 co 10 10 10 

10 D ^ 10 

C4 CO 0> &4 COCOOoOt>U3'^lOlO <OO>OOH^JoqU5l>CDcO ^<N0>HC^G 


; i ; ; i^s i ; 

: : : : : :_2 ^g 
tiftil * 

I > 







lOCO "< 

O IO QQ^J ^.^ 


N 1 ^ COCO OOrH 

t>CO ^O COO 

rH "*' ^P IO O* rH 
CO 3< 

NOO -:Tt ^t- 

t>rH yj'^ 00 00 

cot-^ ^00 co t oq 

COO* ON t-N 
t- CO 

NCO ^q qoq 

rH 10*4 CON 

l> g IO OO 

CO'N CON oo N 

IO < 

<ow <o. ^. w 


rHrH <0CO OlO 

T . 


OrH COO O'O* NrH 0000 

i COrH CO^J ^J^ ,0* 
i t>O* OOO rHN rHrH 
N C-rH rj* 

g'g C 

I COrH rHCO N CD ^< C^ IO CO G "* CO 00 ^J 




^N 00O 

COrH t^TjJ IOIO rHlO l> rH 

cod ON odd cd^ o* 1 ^ 

t- 00 t t- t 

qo ^oq qt^ NIO oco 


oqq Tijq too rnq oqco 

oo toco n<co I>TJ< 

00 O 

l>|> lOCO rHN CON 



^N S^ fil^ ^^ 

I IACO coto t^cq toco 

It* 00 t- IO 

tdd OOrH 

CON CO 1 *! 

OOO ^JJ to 

oo gjco 



. 00^ 

I>N qe^ 


to co 


00 N O 00 rH t> 



rHO rHt^ OCO O t> 
l> 00 CJD O 

COt> t^t- OON t>t*j ^^ ^.^1 "^^ ^^ ^^ 

rHO tf O OCO 
O COrH t> 

NOO rHO N^f COU3 CO 00 

do NO cod did o^ 
t oo oo c t- 

OrH OOO "*O CON 00 1- 

OrH 00^ NCO t>rH rHCO 

OOrH IO rH OO t ^< OO 

CO 00 CO 

t>co 4 oqq coco oqt^ 
S 00* S" 5 N 


i rH CO O 00 

; ^rr ^'d 


IO 00 IO rH CO tO NO CO rH . 
COO COO 0000 t*CN COrH 

t N r* co i 

IOU3 00 tO 


^ N C CO 
tdd OOrH 

55 ^ 

I>O* OrH 

O O ^ CO 

23 cS2 

^ rH 

S3 33 

1 S c 
I I 






^iacb rf oq t>t> coco ocq 
oo> ^fd o^' co *d oico CM'IH cdid "*co 

<0* COCO 00 t> CO<N 0^ 00 CO 10 (N COCO 


Abbe, Cleveland, 367-368 
Absolute humidity, 44 
Absolute temperature scale, 16 

radiant energy, 84 

solar constant, 87 
Acclimatization, 357-358 
Adiabatic processes and stability, 100- 

adiabatic chart, 104-106, (chart) 104, 

atmospheric layers, 116-120 

lapse rates, 106-109 

stability and instability, 109-116 

temperature changes, 100-106 
Adiabatic rate, 101 

saturation, 101 

wet, 101 

Advection, defined, 98 
Aerovane, 36-37 
Agricultural meteorology, 362-365 

climate and crops, 362-363 

droughts, 364-365 

phenology, 365 
Agricultural services of U.S. Weather 

Bureau, 261, 373 
Air (see also Atmosphere) 

anticyclonic circulation, 149 

composition of, 5-8 

conductivity of, 97, 292-293 

conservative properties of, 248-251 

convection in the, 99 

cooling power, health and, 356-357 

cyclonic circulation of, 149 

density, 11 

effect of earth's rotation, 145-147 

humidity and movement of, 355-356 

latitudinal interchange of air, 170-173 

mixing process, 170 

moving (see also Wind) 
polar-equatorial, 154-155 
three forces affecting, 147-148 

pressure of, 24-32 

Air (Cont.): 

stability and instability, 109-116, 
(figs.) 110, 111 

supersaturated, 44 

temperature, 14-24 

upper (see Upper-air) 

vapor pressure, 43 
Air conditioning, health and, 359 
Aircraft weather reconnaissance, 78-79 

to chart path of hurricane, 220 

upper-air observations, 247 
Airfields, fog dispersal, 131 
Air masses: 

analysis of, 174, 247-248 

arctic, 175-176 

classification, 175-176 

climatology, 305 

cold, 176 

continental, 176 

cross-section, 248, (fig.) 249 

daily records of, 305 

equatorial, 175-176 

general circulation of, 157-173 

identifying, 248-251 

latitudinal interchange of air, 170-173 

maritime, 176 
tropical, 181-182 

mixing ratio, 250 

modifications of, 176-178 

nature of, 174-175 

of North America, 176 
characteristics of, 178-183 

polar, 175-176 

continental, 178-180 
maritime, 180 

Rossby diagram, 250-251, (fig.) 251 

source regions, 175 

sources of energy, 172-173 

specific humidity, 250 

structure of, 247-248 

superior, 182-183 

thunderstorms, 225 

tropical, 175-176 
continental, 180-182 
maritime, 181-182 




Air masses (Con*.): 

upper-air, 247-251 

differential analysis, 253 

isentropic analysis, 255, (chart) 256 

warm, 176 

Airway substations, 370 
Albedo, defined, 92 
Aleutian low, 166 
Altimeters, 30-31 

absolute, 31 

pressure, 30, 290 

radio, 31, 290 

Altocumulus clouds, 56, (figs.) 60, 61 
Altostratus, 56, (figs.) 62 

air-masses, 247-248, 260 

canned analyses, 372 

differential, 253 

forecasting and, 243-277 (see also 
Forecasting ) 

isentropic, 255, (chart) 256 

synoptic charts, 243-246 

in the tropics, 208-210, (fig.) 209 

upper-air, 247-259 

mean weather charts, 259 

weather, 243-246 
Anemometers, 35-38 

aerovane, 36-37 

deflection, 35-36 

exposure of, 39 

Gurley electric, 38 

pressure-tube, 36 

Robinson cup, 36, (fig.) 36 
Anemoscopes, 32-33 
Aneroid barometer, 27-29, (fig.) 28 
Angle of incidence, 89-90, (fig.) 90 
Anomalies, from average temperatures, 


Antarctic circle, 160-161, 166 
Anticyclone, 200-201 

Azores, 165 

characteristics of, 200 

cold or shallow, 208 

origin of, 207 

polar-front theory, 202-203 

Pacific, 165 

tracks and velocities, 200-201, (fig.) 

types of, 207-208 

typical paths of, (fig.) 201 

warm or deep, 208 
Anticyclonic circulation, 149 
Appleton layer, of upper air, 119-120, 

(/g.) 117 

Arctic air masses, 175-176 
Asiatic monsoon, 170-172 

Atmosphere (see also Air) 

adiabatic processes and stability, 100- 


composition of, 5-8 
dust, 7-8 

permanent gases, 5-7 
water vapor, 7 
content, weather changes due to, 337- 


cross section, 248 
differential analysis, 253 
elements of weather, 11-13 
meteorological elements, 12 
weather and climate, 12-13 
extent of, 2 

general characteristics, 8-9 
laws of the gases, 9-11 
layers, 116-120, (fig.) 117 
ionized, 119-120 
ozone, 119 
stratosphere and troposphere, 116- 


meteorology defined, 2-5 
pressure of, 11, 24-32 
properties of, 8-11 
stability and instability of air, 109- 


sunspots affect, 351-352 
upper air (see Upper air) 
water vapor, 7, 42 
Auroras, 293-294, (fig.) 294 

types of displays, 294 
Autumnal equinox, 88 
Aviation and the weather, 278-291 
airways observation, 278-279 
all-year round-trip flights across North 

Atlantic, 290 
broadcasts to pilots, 279 
bumpiness, 289 
ceilings and visibility, 73-75 
Flight Advisory Weather Service, 278, 


flying weather, 281-290 
fog, visibility, ceiling, 286-287 
forecasts, 261, 280-281 
GC A ( ground-controlled approach ) , 


hourly sequence, 278 
icing of aircraft, 282-286 

carburetor icing, 285, (fig.) 286 
types of ice deposit, 283-284, 

(figs.) 283, 284 
pressure-pattern flying, 290 
sequence reports, 278, 279-280 
technical landing aids, 287-288 



Aviation and the weather (Con*.): 
thunderstorms as a major hazard, 281- 


turbulence, 288-290 
upper-air observation, 75-80 
VOLSCAN for landing planes, 288 
warm and cold fronts as a cause of 

accidents, 289-290 
weather observations, 278-279 
Aviation weather service, U.S. Weather 

Bureau, 374 
Azores high, 165 


Baguio, 212 (see also Cyclones, tropical) 
Ball lightning, 229, (fig.) 230 

ceiling, 74 

pilot, 76, 168, 247, 281 
Bar, unit of atmospheric pressure, 26 
Barograph, 29, (fig.) 29 

aneroid, 27-29, (fig.) 28 

barographs, 29, (fig.) 29 

barometric tendency, 29 

corrections, 26-27 

Fortin, 24 

lowest observed pressures, 215 

mercurial, 24-27, (figs.) 24, 25 
units of pressure measurement, 26 

reduction of pressure to sea level, 30 

scales compared, 26-27, (fig.) 27 

standard weather observation, 80 

Torricelli, 24 

used for local forecasts, 273 
Barometric pressure: 

depression, 197, (fig.) 197 

low pressure centers, 193-200 
Beaufort wind scale, 35 
Bergeron ice-crystal theory, 131 
Berkofsky, Louis, 272n 
Berry, F. A., 259n 
Bibliography, 379-384 
Bioclimatology, 354 
Biometeorology, 354 
Bjerknes, V., 202-203, 367 
Blair, T. A., 137 
Blizzard, 240 
Bora winds, 153 
Boyle's law, 9, 100 
Braham, R. R. Jr., 222 
Breezes, 150-153 

land, 151-152, (fig.) 151 

mountain, 152-153 

sea, 150-151, (fig.) 150 

valley, 152 

Bumpiness, caused by turbulent move- 
ments of air, 289 
Buys-Ballot's law, 149-150, 198 
Byers, H. R., 222n 

Calorie, definition of, 85 
Campbell-Stokes sunshine recorders, 73 
Canned analyses, used in forecasting, 


balloons, 74 

ceilometer, 75 

definition of, 74 

light, 74 

hazard in aviation, 286-287 

measuring ceiling height, 74, (fig.) 


observation of visibility and, 73-75 
standard weather observation, 79 
Ceilometer, 7 
Celius thermometer, 16 
Cellular circulations, 161, 162 
Centers of action, 162, 257 
Centigrade scale, 16-17 
conversion of Fahrenheit scale to, 388, 


Charles' law of gases, 9-11, 100 

adiabatic, 104-106, (charts) 104, 105 
constant-pressure, 251-255, (figs.) 

254, 258 

five-day time mean chart, 259 
isentropic, 255, 256 
mean weather, 259 
prognostic, 262 
space mean chart, 259 
synoptic, 192, (figs.) 192, 194, 195, 


analysis of, 243-246 
Chinook winds, 238 
Circulation, atmospheric; 
anticyclonic, 200 
cellular, 162 
convectional, 98-99 
cyclonic, 198 
general, 157-173 
index, 255, 257, 259 
January and July of pressure and 

winds, 162-167 

latitudinal interchange of air, 170-173 
secondary, 191-211 
cyclones, 193-200 
high-pressure centers, 200-201 
low-pressure centers, 193*200 



Circulation, atmospheric (Cont.): 
in the tropics, 208-210 
upper-air, 168-170 
winds aloft, 168-170 
yearly averages of pressure, 157-162 
Cirrocumulus clouds, 56, (fig.) 59 
Cirrostratus clouds, 56, (fig.) 60 
Cirrus clouds, 55, (fig.) 59 
Civil Aeronautics Administration, pro- 
vides weather data, 278, 280, 374 

effect of climate on, 359-360 
climate and the beginning of, 360- 


climatic hypothesis of, 359-360 
temperature and, 361-362 

acclimatization, 357-358 
adequate description difficult, 302 
agricultural meteorology, 362-365 
air-mass climatology, 305 
classification system, 328-330, (fig.) 


major divisions, 328-330 
subdivisions, 330 
coastal or littoral, 326 
cold caps, 325 
continental, 325 
crops and, 362-363 
culture and, 359-361 

and the beginnings of civilization, 

climatic hypothesis of civilization, 


influence of extreme climates, 360 
cyclical changes of weather and, 330- 

data, 302-305 

analysis of, 303-305 
kinds of, 303 

punched-card method, 304-305 
definition, 12 
droughts, 364-365 
elements, 302-306 
equatorial zone, 322 
hot belt, 324-325 
humid subtropical, 323-324 
intermediate zones, 324 
major factors influencing, 327 
man's response to, 340-365 
marine, 326 
Mediterranean, 323 
monsoon, 324 
mountain and plateau, 327 
periodic fluctuations, 336-339 
physical, 305-306 

Climate (Cont.): 
polar zones, 324 

precipitation, 315-322 (see also Pre- 
cipitation ) 
related to physical features of earth, 


secular trends, 333-336, (fig.) 334 
solar, 305-306 
steppes, 324 
study of, 302 
subtropical, 323-324 
temperate belts, 325 
temperature, 306-315, (see also Tem- 
perature ) 

theories of climatic changes, 336-339 
atmospheric content, 337-338 
distribution and elevation of land, 

relative position of earth and sun, 


solar radiation, 337 
stability of historic climate, 338- 


trade wind, 323 
tropical zone, 322-323 
weather and, 12-13 
weather cycles, 331-333, (fig.) 332 
zonation of, 322-325 
Climatological Data, 373, 375 
Climatological substations, 320 
air-mass, 305 
science of, 302 
synoptic, 305 
Cloudbursts, 132 
Cloudiness, records of, 67 

altocumulus, 56, (figs.) 60, 61 
altostratus, 56, (figs.) 62 
ceiling, observations of, 73-75 
cirrocumulus, 56 (fig.) 59 
cirrostratus, 56, (fig.) 60 
cirrus, 55, (fig.) 59 
classification, 55-57 
convergence and eddy motions, 135 
cumulonimbus, 57, (figs.) 65-66 
cumulus, 56-57, (fig.) 65 
eddy motion, 135 
fair weather clouds, 57 
hazard in aviation, 283-284, 288 
height and groupings of cloud forms, 


international forms of, 55-66 
nimbostratus, 56, (fig.) 64 
observations, 55-67 
orographic uplift, 134-135 



Clouds (Con*.): 

penetrative convection, 133-134 

precipitation and, 131-135 

records of cloudiness and, 67 

scud, 231 

standard weather observation, 79 

stratocumulus, 56, (fig.) 63 

stratus, 56, (fig.) 64 
Coastal climates, 326 
Cold caps, 325 

Cold fronts, 186-188, (fig.) 187 
Cold wave, 188, 240 

warning service, 261 
Condensation, 122-142 

artificial rain stimulation, 133, 139- 

clouds and precipitation, 131-135 
convergence and eddy motion, 135 
orographic uplift, 134-135 
penetrative convection, 133134 

dew point, 43-44 

above earth's surface, 126-131 
drizzle, 127-128 
fog, 128-131 
haze, 127-128 

forms of precipitation, 135-139 

hygroscopic particles, 127 

latent heat of, 44, 172 

lifting condensation level, 101, 113 

nuclei of, 127 

pressure, 255 

on solid surfaces, 122-126 
dew, 123 
frost, 123-126 
Conduction, 96-97 

air temperatures, 97 

earth temperatures, 97 
Conductivity of air, 292-293 
Conrad, V., 348n 

Conservative properties of air, 248251 
Constant, solar, 85 
Constant-pressure charts, 251-255, (figs.) 

254, 258 

Continental air masses, 176 
Continental climate, 325 
Continental highs of winter, 166 
Continentality, index of, 326 
Controlling the weather, 133, 139-141, 

Convection, 98-99 

in the air, 99 

circulation, 98 

level of free, 113 

in a liquid, 98 

penetrative, 133 

theory of the origin of cyclones, 202 

Convective condensation level, 134 
Convective instability, 223-225 

cause of clouds and precipitation, 135 
intertropical, 210 

Conversion factors and tables, 385-389 
Cooling power of air, health and, 356- 

Co-operative activities of the U.S. 

Weather Bureau, 370, 372-373 
Coriolis force, 145-147 
tropical cyclones, 217 
Corona, 299 

Correlation coefficients, 347-348 
Correlations, weather, 347-348 
Crolls' theory, 337 

climate and, 362-363 
weather and, 363-364 
Crop substations, 370 
Culture and climate, 359-361 
Cumulonimbus clouds, 57, (figs.) 65-66 
Cumulus clouds, 56-57, (fig.) 65 

North Atlantic, 349-350 
North Pacific, 307 
ocean, 308-309 
Peru, 350-351 
wind, 32 
Cyclical changes of weather and climate, 

330-336, (fig.) 332 
Cyclogenesis, 184-185 
Cyclones, 197-200 

characteristics of, 197-198 
cyclonic circulation, 198 
extratropical, 197, 210 

breeding place of, 184 
movement of, 198-200, (fig.) 199 
nature and origin of, 201-208 
convection theory, 202 
movement of air in, 206-207 
occlusion of a wave cyclone, 204- 


polar-front theory, 202-203 
secondary low-pressure centers, 

205-206, (fig.) 206 
vertical cross section, 205-206 
wave theory, 203-204, (fig.) 204 
source regions, 198 

secondary low, 198 
tropical, 212-220 

characteristics, 213-214 
effects of, 218-220 
eye of, 213-214, (fig.) 213 
Florida Keys storm of September 
1935, 214-215 



Cyclones (Cont.)i 

lowest observed pressures, 215 
origin and path, 217-218, (fig.) 216 
regions and times of occurrence, 216 
of Sept 22, 1948, (chart) 219 
warning service, 218, 220 

troughs of low pressure, 197 

typical paths of, (fig.) 199 

V-shaped depressions, 197 
Cyclonic circulation, 149 

Daily forecasts, 262-267, (figs.) 263, 

264, 265 

Daily Weather Map, 375 
Data, climatic, 302-305 

analysis of, 303-305 

punched-card method, 304-305 

kinds of, 303 

angle of incidence and length of, 88- 

degree days of cooling, 359 

degree days of heating, 359 

length of day at each ten degrees of 

latitude, (table) 91 
Degree days of cooling, 359 
Degree days of heating, 359 
Density of air, 11 

barometric, 197, (fig.) 197 

cyclonic, 197 
Dew, formation of, 123 
Dew point, 43-44 

air-masses, 250 

condensation and, (table) 48 

hygrometer, 46 
Diffraction of light, 299-300 
Dines, 33 
Discontinuities, 175 

surface of, 203 
Display substations, 370 
Diurnal variations: 

pressure observations, 31-32 

wind, 40-41 
D layer of atmosphere, 119-120, (fig.) 

Doldrums, 159, 162, 165 
Douglass, A. E., 335 
Drizzle, 127-128 
Dropsondes, use of, 78 
Droughts, 364-365 
Duane, J. E., 214 
in atmosphere, 7-8 

Dust (Cont.): 

falls, 240-241 

nuclei of condensation, 127 

storms, 239, 240-241 
Dynes, 26 



condensation above surface of, 126- 

distance from sun, 87-88 

effect of rotation on winds, 145-147 
Coriolis force, 145-147 

electrical field of, 292 

orbit around the sun, 88-91, (fig.) 89 

temperatures, 97 

weather changes due to relative posi- 
tion of sun and, 337 
Easterlies, polar, 167 
Easterly waves, 210 
E layer of atmosphere, 119-120, (fig.) 

Electrical phenomena, 292-294 

auroras, 293-294, (fig.) 294 

conductivity of the air, 292-293 

electrical field of the earth, 292 

lightning, 229-230 
Electric charge, in thunderstorms, 227- 

229, (fig.) 228 

Electromagnetic spectrum, 83, (fig.) 83 
Electromagnetic waves, 82-83 
Electronic calculator, used in weather 

forecasting, 272 

radiant, 82-84 

sources of, in air, 172-173 
Equator, movement across the, 172 
Equatorial air masses, 175-176 
Equatorial belt of low pressure, 159 
Equatorial climatic zone, 322 
Equatorial-polar air movements, 154- 


Equatorial trough, 208, 209 
Equinox, vernal and autumnal, 88 
Equivalent potential temperatures, 103- 

upper-air masses, 250-251 

amount of, 52-53 

definition of, 42 

fogs, 130 

measurement of, 53-55 

observations, 51-55 

salinity and, 53 

station, (fig.) 54 

wind movements and, 53 




anemometer, 39 

rain gages, 69 

thermometers, 20-21 

wind vanes, 39 

Extended forecasts, 259, 268-271 
Extrapolation method, of preparing fore- 
casts, 269 

of precipitation, 321-322 

of temperature, 312 

Fahrenheit thermometer, 17 

conversion of centigrade scale to Fahr- 
enheit, 388, 389 
Fallwinds, 153 
Fawbush, E. J., 236 

Fire-weather forecast and warning serv- 
ice, 261, 374 

First-order stations, 370, 371-372 
Five-day time mean chart, 259, 268-269 
F layers of atmosphere, 119-120, (fig.) 


Flood, forecast and warning service, 261 
Flora, S. D., 235 
Florida Keys storm of September 1935, 


Foehn wind, 237, (fig.) 239 
Fogs, 127-131 

advection, 129-130 
classification of, 128 
cost and dispersal, 131 
evaporation, 130 
frontal, 130 
ground, 128-129 
hazard to aviation, 286 
high-inversion, 128, 129 
radiation, 128-129 
smog, 128 
steam, 130 
upslope, 130-131 
Forecasting, weather: 

analysis of synoptic charts, 192-193, 
243-246, (figs.) 192, 194, 195, 
estimating movement of highs and 

lows, 244-245 
estimating the resulting weather, 


placing of fronts, 246 
aviation forecasts, 261, 280-281 
canned analyses, 372 
charting of data, 260 

Forecasting, weather (Con*.): 

daily forecasts, 262-267, (figs.) 263, 
264, 265 

definition, 243 

electric calculator for, 272 

extended, 259, 268-271 
method of extrapolation, 269 
method of five-day mean, 268-269 
method of weather types, 269-271 

five-day time mean chart, 259 

George method, 261, 266 

by government agencies, 261, 370, 

hurricanes, 220 

jet stream, 169 

local forecasts, 273-277 
single-station, 275-276 
statistical indications, 273-275 

long-range forecasts, 271-272 

numerical weather prediction, 272 

polar sequence or polar cycle, 271 

private weather forecasters, 261, 377 

process of, 259-260 

prognostic charts, 262 

radar storm detection, 267-268 

seasonal, 271, 352-354 
hope of, 353-354 

nature of atmospheric responses, 

by the Severe Local Storm Center, 372 

short-range forecasts, 267 

space mean chart, 259 

special forecasts, 372 

thunderstorms, 226, 229 

tornadoes, 235-236 

tropical cyclones, 220 

types of, 260-261, 371-372 

U.S. Weather Bureau, 261, 370, 371- 

upper-air analysis, 75-80, 247-259 
analysis of air masses, 174, 247-248 
circulation index, 255-259 
constant-pressure (pressure con- 
tour) charts, 251-255 
cross sections, 248, (fig.) 249 
identifying air masses, 248-251 
isentropic analysis, 255 
mean weather charts, 259 

weather, 259-270 

weather control, 276-277 

weather lore, 273-277 

weather proverbs, 276 

wind-barometer indications, 273 
Forest fire-warning service, 374 
Fortin barometers, 24 
Franklin, Benjamin, 227 



Frequency polygon, 341-342, (fig.) 342 
Frontogenesis, 184 
Frontolysis, 184 

characteristics of, 183-189 

cold, 186-188, (fig.) 187 
secondary, 188 

cyclogenesis, 184-185 

definition of, 183 

forecasting, 246 

formation of, 183-189 

frontal fogs, 130 

frontogenesis, 184 

frontolysis, 184 

hazard to aviation, 289-290 

intertropical, 210 

nature of, 175 

occluded, 189 

polar, 167 

other types of, 189 

squall lines, 188-189 
pre-cold-frontal, 188 

stationary, 189 

thunderstorms, 224-225 

upper, 189 

warm, 185-186, (fig.) 185 
Frost, 123-126 

black, 123 

forecasts of, 370, 373 

hoarfrost, 123 

killing, 123 

protection against, 124-126, (figs.) 

125, 126 
heaters, 124-125 
wind machines, 126 
Frost Charts, 375 

Fruit-frost service of Weather Bureau, 
370, 373 


rain, 67-69, (figs.) 68, 69 
8-inch, 67-68, (fig.) 68 
exposure of, 69 
Fergusson weighing rain and snow, 

68-69, (fig.) 69 
recording, 68 

condensation, 44 
laws of, 9-11 
Boyle's law, 9 
laws of Charles and Gay-Lussac, 9- 


permanent, in abnosphere, 5-7 
water vapor, 7, 42 
Gay-Lussac, law of gases, 9-11, 100 

General circulation, 157-173 
Geographic distribution: 

of precipitation, (fig.) 321 

solar radiation, 92-93, (fig.) 93 

temperature, 306-315 

thunderstorms, 225-227, (chart) 226 

tropical cyclones, 216 
George, Joseph J., 261, 266 
Geostropic winds, 148, 254-255 
Glaciers, 336 
Glaze, 139, (fig.) 140 

pressure, 143-145 

winds, 148-149 

correction for, 26 

winds, 153 

Great Lakes, weather service for, 368 
Gunn, Ross, 227n 
Gurley electric anemometer, 38 


Hagenguth, J. H., 229n 
Haggard, W. H., 259n 
Hail, 137-139 

formation of, 138-139, (fig.) 138 

soft, 139 

Hales, W. B., 153n 
Hall, Howard E., 290 

parhelion, 299 

solar and lunar, 298-299, (figs.) 299, 

Haze, 127-128 

dry, 128 

moist, 128 

optical, 296 

acclimatization, 357-358 

air conditioning, 359 

air-movement and, 355-356 

cooling power of air and, 356-357 

humidity and, 355-356 

ideal weather, 358 

sunshine and, 358 

temperature and, 354-355 

weather and, 354-359 

conduction, 96-97 

convection, 98-99 

definition of, 14 

degree days of heating, 359 


of condensation, 44 
of sublimation, 42-43 
of vaporization, 42 



Heat (Con*.): 
rays, 83 

temperature and, 14 
waves, 84 

Heater, orchard, 126 

of atmosphere, 252n 
measuring height of clouds, 57 
pressure varies with, 29-31 
Henry, Joseph, 367 
Hertzian electric waves, 83 
Highs, barometric (see Anticyclones) 
Hoarfrost, 123 
Hill, Leonard, 356 
Holzman, B., 53 
Horse latitudes, 159 

Horticultural services, U.S. Weather Bu- 
reau, 261, 323 
Hot belt, 324-325 
"Hot winds" of the Great Plains, 355- 


Houghton, Henry G., 92-93 
Hourly sequences, 278-280 
Hoyt, John R., 287n 
absolute, 44 
dew point and condensation, 43-44 

(table) 48 
health and, 355-356 
measurement of, 46-51 
mixing ratio, 45 
observations, 42-51 
records, 51, (fig.) 52 
relative, 45, (table) 49 
specific, 44-45, 250 
supersaturated air, 44 
vapor pressure and saturated vapor, 

43 (table) 43 
Humphreys, W. J., 155, 292n 
Hurricanes, 212 (see also Cyclones, trop- 

aircraft reconnaissance, 220 
Florida Keys storm of September, 

1935, 214-215 
forecasting, 220 
tracking and locating, 78 
warning service, 218, 220, 261 
Hydrologic services, U.S. Weather Bu- 
reau, 374 

Hydrometer, definition of, 126-127 
Hygrograph, hair, 50, (fig.) 52 
dew-point, 46 
electric, 51 
hair, 50 
Hygroscopic particles, 127 

Ice, weather and polar, 350 
Iceland low, 166 
Ice storms, 139, (fig.) 140 
Icing of aircraft, 282-286 

types of ice deposit, 283-284, (figs.) 

283, 284 

circulation, 255, 257, 259 

of continentality, 326 

amount received at a fixed location, 

defined, 85 

effect of angle of incidence on, 89- 
90, (fig.) 90 

reflection of, 94-96 
Instability of atmospheric conditions: 

auto-con vective, 112 

conditional, 112-114, (fig.) 113 

convective, 114-115 

level of free convection, 113 

potential, 115 
Instruments, weather, 366-376 

shelter for, 20 

Intertropical convergence, 210 
Intertropical front, 210 
Inversion of temperature, 107 
Ionized layers, 119-120 
Ionosphere, 120 
Isallobars, 244 

Isentropic analysis, 255, (chart) 256 
Isobaric surfaces, 144-145 

high-pressure centers, 200 

low-pressure centers, 193 

pressure gradients and, 143-145, 
(figs.) 144, 145 

surface winds in relation to, 149, (fig. ) 


Isohyets, indicate precipitation on weath- 
er maps, 315 
Isothermal lines, 307 
Isothermal region, 116 (see also Strato- 

Isotherms, 307, (figs.) 308, 310, 311, 
313, 314 


Jet stream, 169-170 
Johnson, D. H., 169n 
Jones, H. L., 235-236 
Jungle, 322-323 



Katabatic winds, 153 
Kata-thermometers, 356 
Kennelly-Heaviside layer, 120 
Kincer, J. B., 334-335 
Kite-flying stations, 75 
Koppen, Wladimir, 328 


breeze, 151-152, (fig.) 151 

effect of solar radiation on, 94-95 

weather changes due to distribution 

and elevation of, 338 
Lapham, Increase A., 368 
Lapse rates, 106-109 

examples, (fig.) 108 

stability and instability, 109-116, 
(figs.) 110, 111 

thunderstorms, 223 

turbulence in relation to, 115-116 

variability, 107-109 
Latent heat: 

of condensation, 44 

of sublimation, 42-43 

of vaporization, 42 

Latitudinal interchange of air, 170-173 
Layers, atmospheric, 116-120, (fig.) 117 
Level of free convection, 113 
Lifting condensation level, 101, 113 

diffraction, 299-300 

refraction, 295-296 

scattering, 299-300 

speed of, 83 

ball, 229, (fig.) 230 

nature of, 229 

protection against, 229-230 

rods, use of, 229-230 

sheet, 229 

Liquid, convection in a, 98 
Idtttjfral climates, 326 
Locallforecasts, 273-277 
Long-range forecasts, 271-272 
Low-pressure centers, 193-200 

secondary, 198, 205 
Ludlum, David M., 321 
Lunar halos, 298-299, (figs.) 299, 300 


McDonald, W. F., 215 
"Mackerel sky," 58 

Maps, weather, 191-193 

atmospheric conditions in the tropics, 
208-210, (fig.) 209 

facsimile, 375-376, (fig.) 376 

high-pressure centers, 200-201 

isallobars, 244 

isobars, 143-144, (figs.) 144, 145 

isothermal lines, 307 

low-pressure centers, 193-200 

making, 192-193 

pressure change chart, 244-245 

reproducing, 375-376, (fig.) 376 

synoptic charts, 192, (figs.) 192, 194, 

195, 196 

analysis of, 243-246 
estimating movement of lows and 
highs, 244-245 

temperature, 307 

value of, 193 

wind direction and velocity, 34, 35 
March of temperature: 

annual, 23 

daily, 21 

Marine climates, 326 
Marine meteorological service, U.S. 

Weather Bureau, 373-374 
Marvin, C. F., 348n 
Masses of air (see Air-masses) 
Mean weather charts, 259 

five-day, 259 
Median, 341 

Mediterranean climate, 323 
Mercurial barometers, 24-27 
Meteorograph, records wind direction, 32 

agricultural, 362-365 

brief history of, 366-367 

early organized weather observa- 
tions, 367-369 

defined, 2-5 

development of a weather service, 367- 

elements, 12 

professional training, 377 

purpose of, 157 

science of climatology and, 302 
Microseisms, 220 
Middleton, W. E. K., 15n 
Military weather services, 375-377 
Miller, R. C., 236 
Millibars, 26 

converting to inches, 26, (table) 386 
Mirages, 296-297, (fig.) 297 
Mistral wind, 153 
Mixing ratio, 45 
Mode, 341 



Monsoon climate, 324 
Monsoons, 153-154 

Asiatic, 170-172 
Monthly Weather Review, 375 
Moon, halos, 298-299 

breezes, 152-153 

climates, 327 
Mudge, Robert W., 284n 
Myer, Albert J., 368 


Namias, Jerome, 170n, 259 
Newell, Homer E. Jr., 79 
Nimbostratus clouds, 56, (fig.) 64 

atmosphere, 26 

climate, 13 

defined, 13 

precipitation, 315-317 

values, 13 

weather, 13 
North America, air masses, 176 

characteristics of, 178-183 
Northern lights, 293-294 
Norther wind, 239-24 



airways, 278-279 

balloon, 76, 168 

brief history of early, 367-369 

ceiling and visibility, 73-75 

clouds, 55-67 

evaporation, 51-55 

humidity, 42-51 

kite, 75 

making standard weather, 79-80 

precipitation, 67-71 

pressure, atmospheric, 24-32 

sunshine, 71-73 

temperature, 14-24 

upper-air, 75-80 

visibility, 73-75 

weather, standard, 79-80 

Weather Bureau, 370-375 
Occluded fronts, 189 
Occlusion, 189 

of a wave cyclone, 204-205 
Ocean and lake forecast and warning 

service, 261 

currents, 308-309 

effects of northeast trades, 349-350 

Oceans (Cont.): 

Peru Current and Peruvian rainfall, 


polar ice and weather, 350 
some relations between weather and 

temperature of, 348-349 
weather and, 348-351 
Optical phenomena, 294-301 
corona, 299 

diffraction and scattering, 299-300 
mirage, 296-297, (fig.) 297 
optical haze, 296 
parhelion, 299 
rainbows, 297-298 
refraction, 295, (fig.) 295 

atmospheric, 295-296 
solar and lunar halos, 298-299, (figs.) 

299, 300 
twilight, 300-^301 
Orchards, heaters for, 126 
Oscillographs, for locating thunder- 
storms, 229 
Ozone layer, 119 

Pacific anticyclone, 165 

Pacific high, 165 

Pampero, wind in South America, 240 

Parhelion, 299 

Peru Current and Peruvian rainfall, 350- 

Petterssen, Sverre, 126n, 262n, 267 

Phenology, 365 

Physical climate, 30,5-306 

Pibals, 76 

Pilot-balloon observations, 76, 168, 247, 

Plateau climates, 327 

Polar air masses, 175-176, 178-179 
continental, in summer, 179 
continental, in winter, 178-179 
maritime, 179-180 

Polar caps of high pressure, 160-Jd| 

Polar climatic zones, 324 '7 1 ''' 

Polar easterlies, 167 

Polar-equatorial air movements, 154-155 

Polar front, 167 

Polar-front theory, origin of cyclones and 
anticyclones, 202-203 

Polar ice, 350 

Polar low pressure, 160 

Polar sequence or polar cycle, 271 

Pollak, L. W., 348n 

Polygon, frequency, 341-342, (fig.) 342 

Potential instability, 115 



Potential temperature, 103 

accumulated sums of departure, 333 

annual number of rainy days, 318- 
319, (fig.) 320 

areas of light and heavy, 319-321 

artificial rain stimulation, 133, 139- 
141, 276-277 

Bergeron ice*crystal theory, 131 

causes of, 131-135 

climatic data, 302-306 

cloudbursts, 132 

clouds and, 131-135 

distribution of, (fig.) 321 

excessive, 70-71 

forms of, 135-139 

frequency polygon, 341-342, (fig. ) 
342, 343 

general distribution of, 315-322 

glaze, 139, (fig.) 140 

hail, 137-139 

formation of, 138-139 

ice storm, 139, (fig.) 140 

in Iowa, 1873-1955, (fig.) 332 

isohyets used to indicate, on weather 
maps, 315 

mean monthly and annual tempera- 
tures and, (tables) 390-396 

normal annual, 315-317, (fig.) 316 

observations, 67-71 

penetrative convection, 133-134 

Peruvian rainfall, 350-351 

rain gages, 67-69 
exposure of, 69 

Fergusson weighing rain and snow, 
68-69, (fig.) 69 

records, 70-71 

seasonal variations, 317-318, (fig.) 

sleet, 139 

snow, 137 

snow grains, 139 

some extremes of, 321-322 

standard weather observation, 80 

tables, 390-396 

unusual rainfalls in the U.S. (fig.) 321 

weather cycles, 331-333, (fig.) 332 

weather maps, 315 
Pre-cold-frontal squall lines, 188 
Pressure, atmospheric: 

Aleutian and Iceland lows, 166 

altimeters, 30-31 

aneroid barometer, 27-29, (/ig.) 28 

Azores high or Azores anticyclone, 165 

barographs, 29 

Buys-Ballot's law, 149-150 

Pressure, atmospheric (Cont.): 
centers of action, 162 
circulation index, 255, 257, 259 
circulation zones and cells, 161-162 
condensation, 255 
constant-pressure chart, 251-255, 

(figs.) 254, 258 
continental highs of winter, 166 
contour chart, 251-255, (figs.) 254, 


density of air, 11 
diurnal variations, 31-32 
equatorial belt of low, 159 
general circulation, 157-173 
gradients, 143-145 
high-pressure belts, 165 
interrelation of temperature, wind, 

and, 143-156 
isobars and pressure gradients, 143- 

145, (figs.) 144, 145 
January mean, 162-167, (/ig.) 163 
July mean, 162-167, (fig.) 163 
lowest observed, 215 
low-pressure centers, 193-200 
mean annual sea-level, (fig. ) 158 
mercurial barometer, 24-27, (figs.) 

24, 25 

movement across the equator, 172 
nature and origin of highs and lows, 


observations of, 24-32 
Pacific high or Pacific anticyclone, 165 
polar caps of high pressure, 160-161 
polar low pressure, 160 
pressure change chart, 244-245 
pressure-tube anemometer, 36 
reduction to sea level, 30 
results of observations, 31-32 
saturation, 255 
semipermanent centers, 162 
standard weather observation, 80 
subtropical high-pressure belts, 159- 


in the tropics, 208-210 
troughs of low, 197 
units of pressure measurement, 26 
upper-air, 168-170 
vapor pressure and saturated vapor, 43, 

(table) 48 

vapor pressure of air, 51-53 
vapor pressure of water surfaces, 51 
variations of pressure with height, 29- 

wind belts and, of the world, (fig.) 

yearly averages, 157-162 



Pressure-pattern flying, 290 
Prevailing westerlies, 166-167 
Private weather services, 261, 377 
Proverbs, weather, 276 
Psychrometers, 46-47, (fig.) 47 

aspiration, 47 

sling, 47 

telepsychrometer, 50 

whirled, 46, (fig.) 47 
Psychrometric Tables of the United 

States Weather Bureau, 50 
Publications, U.S. Weather Bureau, 374- 

375, 383-384 
Punched-card method, for climatological 

studies, 304-305 
Pyrheliometer, 86-87, (fig.) 86 


Radar storm detection, 220, 226, 267- 

Radiant energy, 82-84 

characteristics of, 83-84 

transmission, absorption and reflection, 


definition, 82-83 

solar, 82-89 (see also Solar radiation) 

altimeter, 290 

direction recorders, for locating thun- 
derstorms, 229 

ionized layers, 119-120 

reflections from ionosphere, 119-120 
Radiosondes, 76, (fig.) 77, 168, 247, 
248, 255, 275, 281 

artificial stimulation, 133, 139-141 
seeding operation, 140, (fig.) 141 

formation of, 135-136 

gages, 67-69, (figs.) 68, 69 
Rainbows, 297-298 
Rainfall (see Precipitation) 
Range of temperature, annual and daily, 
23-24, 312, (figs.) 23, 313, 314 
Raobs, 76 

Rawin observations, 247, 281 
Rawinsondes, 78, 168 
Recorders, sunshine, 71-73 
Records, weather: 

brief history of, 367 

clouds and cloudiness, 67 

evaporation, 53-^54 

humidity, 51 

precipitation, 70-71 

sunshine, 73 

upper-air, 75 

Reduction of pressure to sea level, 30 

diffused, 84 

radiant energy, 84 

solar constant, 87 
Refraction, 295 

atmospheric, 295, 296 

twinkling of stars, 296 
Register, triple, 32 
Relative humidity, 45, (table) 49 
Research, meteorological, 371 

joint sponsorship of projects, 377 
U.S. Weather Bureau, 374-375 
Richardson, L, F., 272 
Riehl, Herbert, 169n, 210, 217n 
River and flood forecast and warning 

service, 261, 374 
River substations,, 370 
Robinson cup anemometer, 36, (fig.) 36 
Rocket-sonde research, 79 
Rossby diagram, 250-251, (fig.) 251 

St. Elmo's fire, 293 
Saturated vapor, 43 
Saturation, 43 

adiabatic rate, 101 

pressure, 255 
Savannas, 323 
Scales, Beaufort wind, 35 
Scattering of light, 299-300 
Schonland, B. F. J., 292n 
Sea breezes, 150-151, (fig.) 150 
Sea level, reduction of pressure to, 30 
Seasonal forecasting, 352-353 
Secondary circulation, 191-211 
Secondary low, 198 
Secular trends, climate and weather, 

333-336, (fig.) 334 
Sequence reports, 278-280 
"Sferics" equipment, 268 
Shaw, Sir Napier, 116, 173, 207 
Short-range forecasts, 267 
Simoons, 355-356 
Simpson, R. H., 220 
Sirocco, 238-239 
Sleet, 139 
Smithsonian Meteorological Tables, 30, 

46, 47 

Smog, definition, 128 

blizzard, 240 

crystals, 137, (fig.) 136 

grains, 139 

granular, 139 



Snoifr (Cont.)i 

measurement of, 69-70 

Fergusson weighing gage, 68-69, 

(fig.) 69 

reflects solar radiation, 94 
Snowfall substations, 370, 374 
Solar climate, 305-306 
Solar constant, 85 
Solar radiation, 82-99 
conduction, 96-97 
convection, 98-99 
direct effects of, 91-96 
on air, 92-93 
on land surfaces, 94-95 
on water surfaces, 95-96 
distance from earth to sun, 87-91 
distribution of, 92-93, (fig.) 93 
incoming, 84-87 
insolation, 85 

amount received at a fixed location, 

length of day and angle of incidence, 


measurements, 86-87 
radiant energy, 82-84 
solar constant, 85 

atmospheric absorption and reflec- 
tion, 87 

turbulent atmosphere and, 351-352 
weather changes influenced by, 337 
Solstice, summer and winter, 88 
Space (see Upper-air) 
Spectrum, electromagnetic, 83, (fig.) 83 
Spells, weather, 344 
hot, 344 
rainy, 344 

Squall lines, 188-189, 204 
Squalls, wind, 41 
Stability and instability, atmospheric, 

109-116, (figs.) 110, 111 
conditional instability, 112-114, (fig.) 


convective instability, 114115 
subsidence, 115 
turbulence in relation to lapse rates, 


Starrett, L. G., 236n 
Stars, twinkling is a refraction phenom- 
ena, 296 

evaporation, 53-54 
first-order, 370, 371-372 
pressure, 80 

radiosonde and rawin, 78 
reporting, 260, 370 
Weather Bureau, 370 

Steppe climate, 324 
Stone, R. G., 357n 

dust, 239, 240-241 

hurricanes, 212-220 

radar storm detection, 267-268 

thunderstorms, 221-230 

tornadoes, 230-237 

tropical cyclones, 212-220 
Stormy westerlies, 167 
Stratocumulus clouds, 56, (fig.) 63 
Stratosphere, 116, 120, (fig.) 117 

temperatures in the lower, 118 
Stratus clouds, 56, (fig.) 64 
Streamlines, 208-209 

latent heat of, 42-43 

process of, 42 

Subsidence of a layer of stable air, 115 
Subtropical climatic zones, 323-324 
Subtropical high-pressure belts, 159-160 
Subtropical ridges, 208, 209 

causes convective movements of air, 

distance of earth from, 87-88 

dogs, 299 

earth's orbit around, 88-91, (fig.) 89 

halos, 298-299 

weather and the, 351-354 

weather changes due to relative posi- 
tion of earth and, 337 

important for health, 358 

observations, 71-73 

recorders, 71-73 
Campbell-Stokes, 73 
electric, 71-73, (fig.) 72 

records, 73 

switch, 71, (fig.) 72 
Sunspots, 293 

weather and, 351 
Superior air masses, 182-183 
Surfaces : 

condensation on solid, 122-126 

of discontinuity, 203 

isentropic, 255 

isobaric, 144, (figs.) 144, 145 

land, effect of solar radiation on, 94 

water, effect of solar radiation on, 95- 

winds, 149 

Sverdrup, H. U., 351n 
Synoptic charts, 192-193, (fig.) 194, 
195, 196 



Synoptic charts (Con*.): 
analysis of, 243-246 
constant-pressure chart, 
(figs.) 254, 258 

Synoptic climatology, 305 



conversion factors and, 385-389 

dew point, 48 

length of day at each ten degrees of 
latitude, 91 

mean precipitation, 390-396 

mean temperature, 390-396 

psychrometric, 50 

relative humidity, 49 

vapor pressure, 48 
Telepsychrometer, 50 
Television, meteorologists on, 378 
Temperate belts, 325 

adiabatic changes, 100-106 

adiabatic chart, 104-106, (charts) 
104, 105 

air conditioning, 359 

annual and daily ranges, 23-24, 312, 
(figs.) 23, 313, 314 

annual march, 23-24, (chart) 23 

anomalies or departures from average, 

civilization and, 361-362 

climatic data, 302-306 

correction, 26 

daily march of, 21, 23-24, 312, (figs.) 
23, 313, 314 

distribution of, 306-315 

equivalent potential, 103-104, 248 

extremes of, 312 

health and, 354-355 

instrument shelter, 20-21 

interrelation of wind, pressure and, 

inversion of, 107 

January mean, 309-312, (fig.) 310 

July mean, 309-312, (fig.) 311 

lapse rates, 106-109 
variability of, 107-109 

mean, 21 

mean annual sea-level, (fig.) 308 

mean monthly and annual precipita- 
tion and, (tables) 390-396 

mean virtual, 253 

nature of heat and, 14 

Temperature (Con*.): 

normal yearly, 307-309, (fig.) 308 

observing, 14-24 

obtaining, 19-21 

ocean currents affect, 308-309 

potential, 103 

isentropic analysis, 255, (fig.) 256 
wet-bulb, 250 

relationship between ocean currents 
and weather, 348-349 

retardation or lag, 21 

thermographs, 18-19, (fig.) 19 
record, (fig.) 22 

thermometers, 14-19 

in the upper atmosphere, 120 

use of observations, 21-24 

wet-bulb, 46-47 

wet-bulb potential, 250 
Tepper, Morris, 236 
Thermals, rising currents of unstable air, 

Thermographs/ 18-19 

record, (fig.) 22 
Thermometers : 

absolute scale, 16 

Bourdon, 15 

Celsius, 16 

deformation, 15 

dry-bulb, 46-47 

electrical, 15-16 

Fahrenheit, 16-17 

formulas for converting from one scale 
to another, 16, 388, 389 

instrument shelters, 20-21 

kata-thermometer, 356 

liquid-in-glass, 15 

liquid-in-metal, 15 

marking degrees on, 15-16 

maximum, 17-18 

minimum, 18 

thermographs, 18-19 

wet-bulb, 46 

Thompson, Philip D., 272 
Thornthwaite, C. W., 53 
Thunderstorms, 221-230 

convective, 223-225 

description of local, 221 

detecting, 268 

development of convective instability, 

electric charge, 227-229, (fig.) 228 

geographic distribution, 225-227, 
(chart) 226 

hail, 138 

heat, 224 
artificial, 224 



Thunderstorms (Cont.): 

nature of, 229 
protection against, 229-230 

major hazard of aviation, 281-282 

structure, 223 

cumulus stage, 223 
dispersal stage, 223 
mature stage, 223 

types, 225 
air-mass, 225 
frontal, 225 

violent movements in, 221-223, (fig.) 

Tornadoes, 230-237 

characteristics, 230-234, (figs.) 231, 
232, 233 

destructive forces in, 234 

detecting, 268 

lapse rates, 112 

place and time of occurrence, 234- 

warning, 235-236 

waterspouts, 236-237, (fig.) 237 
Trade wind climates, 323 
Trade winds, 165-166 

antitrades, 168-169 

easterly, 209 

easterly waves, 210 

northeast, effects of, 349 
Transfer of heat: 

conduction, 96-97 

convection, 98-99 

radiation, 84 
Trewartha, Glen T., 328 
Triple register, 32 
Tropical air masses, 175-176, 180-181 

continental, 180-182 
.Tropical climatic zone, 322-323 

atmospheric circulation, 208-210 

pressure, winds and weather, 208-210 

weather analysis, 208-210 
Tropopause, height of, 118 
Troposphere, 116-117, (fig.) 117 

wind and pressure pattern, 168 
Trough of low pressure, 197 
"Turbulence, atmospheric, 38-39 

hazard in aviation, 288-290 

in mountains, 288-289 

upper-air, 39 
Twilight, 300-301 

astronomical, 301 

civil, 301 

Twinkling of stars, 296 
Twister, 230 (see also Tornadoes) 

Typhoons, 212 (see also Cyclones, trop- 

' ical) 
locating and tracking, 78 


Ultraviolet radiation: 

ozone layer absorbs, 119 
United States: 
Air Force: 

weather reconnaissance, 7879 
weather services, 78-79, 261, 280, 


Signal Service, 368 
weather services, 261, 375 
Commerce Department, 369 
Government Land Office, 367 

Aerology Section, 376 
weather services, 261, 280, 375-377 
Patent Office, 367 
Smithsonian Institution, 367 
Weather Bureau: 
activities of, 370-375 
agricultural and horticultural serv- 
ices of, 373 
airway meteorological service, 239, 


central office, 369-370 
climatological service, 372 
daily records of air-masses, 305 
establishment of, 368-369 
field organization, 370 
fire-weather warning service, 374 
Flight Advisory Weather Service, 

278, 280 
forecasting services, 261, 272, 371- 


future outlook for the weather serv- 
ices, 377-378 
hydrologic services, 374 
kite-flying stations, 75 
marine meteorological service, 373- 


present organization, 369-370 
Psychrometric Tables, 50 
in publications, 374-375, 383-384 
radiosonde and rawin stations, 78 
records of clouds and cloudiness, 67 
research and publications, 272, 374- 


tornado-warning service, 236 
Weather Glossary, 126n 



Upper air: 

aircraft weather reconnaissance; 78- 

analysis, 247-259 

constant-pressure (pressure contour) 
charts, 251-255, (figs.) 254, 258 
plotting soundings, 251 
atmospheric layers, 116-120, (fig.) 

observations, 75-80 

at selected stations, 371 
obtaining data, 75-76 
pilot balloons, 76 
radiosondes, 76, (fig.) 77 
rawinsondes, 78 
rocket-sonde research, 79 
temperature, 120 
winds, 168-170 
Upper front, 189 


Vanes, wind, 32 

exposure of, 39 
Vapor pressure, 43 

of air, 51-53 

saturated vapor and, 43 

of water surfaces, 51 
Vaporization, latent heat of, 42 

of light, 83 

wind, 34-35 

effect of altitude, 39 
Vernal equinox, 88 

definition of, 73 

for flying safety, 286-287 

measuring, 73 

observations of, 73-75 

standard weather observation, 79 
VOLSCAN for landing planes, 288 
V-shaped depressions, 198 


Ward, R. DeC., 360 
Warning services: 

fire-weather, 374 

forecasting and, 261, 370, 371-372 

hurricanes, 218, 220, 261 

tornadoes, 235-236 

tropical cyclones, 218, 220 

U.S. Weather Bureau, 370, 371-372 
Water surfaces: 

effect of solar radiation on, 95-96 

vapor pressure of, 51 
Waterspouts, 236-237, (fig.) 237 

Water vapor, in atmosphere, 5, 7, 42 

condensation of, 43 

cold, 188 

easterly, 210 

electromagnetic, 82-83 

heat, 84 

Hertzian electric, 83 
Wave theory, origin of cyclones, 203- 

204, (fig.) 204 
WBAN (Weather Bureau, Air Force, 

Navy), 377 

agricultural meteorology, 362-365 

aircraft weather reconnaissance, 78-79 

analysis and forecasting, 243-277 (see 
also Forecasting) 

aviation and, 278-291 

barometric tendency, 2 

climate differs from, 12 

climatic elements, 302- 

combination ^of element 

correlations, 347-348 
coefficients, 347-348 

crops and, 362-363 

cyclical changes of climate and, 330- 

336, (fig.) 332 
secular trends, 333-336, (fig.) 334 

definition, 12 

departures of mean monthly tempera- 
tures from normal at St. Paul, 
Minn., 345 

droughts, 364-365 

elements of, 11-13 

combinations of, 340-346 

forecasting, 259-270 (see also Fore- 

frequency polygon, 341-342, (fig.) 342 

hazards in flying, 278-291 

health and, 354-359 

ideal, 358 

map (see Maps, weather) 

mean, 341 

mean charts, 259 

median,, 341 

mode, 341 

normal, 13 

oceans and, 348-351 

persistence of, 343-346 

polar ice and, 350 

proverbs, 276 

seasonal forecasting, 352-354 

secular trends, 333-336, (fig,) 334 


military, 375-377 
private, 377, 378 



Weather (Con*.): 

spells, 344 

standard observation, 79-80 

sun and, 351-352 

sunspots, 351 

theories of climatic changes, 336-339 

in the tropics, 208-210 

types, 344-346 

variability of, 340-346 

monthly and annual, 341343 

world- wide relationships, 340-365 
Weather Bureau (see United States, 

Weather Bureau) 

Weather Bureau Research Papers, 375 
Weather-type method, of forecasting, 

Weekly Weather and Crop Bulletin, 373, 


Weeks, Sinclair, 375n 
West Wind Drift, 307 

prevailing, 166-167 

stormy, 167 

Wet-bulb potential temperature, 250 
Whirlwinds, 237 

aloft, 168-170 

anemometers, 3538 
exposure of, 39 

anemoscopes, 32-33 

annual variations, 40 

anticyclonic circulation, 149 

antitrades, 168-169 

Asiatic monsoon, 170172 

Beaufort scale, 35 

blizzard, 240 

bora, 153 

breezes, 150-153 

Buys-Ballot's law, 149-150 

chinooks, 238 

currents, 32 

cyclonic circulation of, 149 

direction, 32-34 

January mean, 162-167, (fig.) 163 
July mean, 162-167, (fig.) 163 

diurnal variations, 40-41 

due to local temperature differences, 

effect of altitude, 39 

faUwinds, 153 

foehn, 237, (fig) 239 

general circulation, 157-173 

geostrophic, 148, 254-255 

gradient, 148-149 

gravity, 153 

gustiness of, 38-39 

"hot winds," 355-356 

Winds (Cant.): 

indicating on weather map, 34-35 

instruments, 32-34 

interrelation of temperature, pressure 

and, 143-156 
irregular variations, 41 
January mean, 162-167, (fig.) 163 
jet stream, 169-170 
July mean, 162-167, (fig.) 163 
katabatic, 153 

latitudinal interchange of air, 170-173 
machines for preventing frost, 126 
measuring, knots, 35n 
meteorograph, 32 
mistral, 153 
monsoons, 153154 
northeast trades, 165 
northers, 239-240 
observations, 3241 

results of, 40-41 
pampero, 240 
polar easterlies, 167 
polar-equatorial air movements, 154 


prevailing westerlies, 166167 
rose, (fe. ) 40 
shaft, 34 

simoons, 355356 
sirocco, 238-239 
southeast trades, 165-166 
special, 237-241 
speed, 34-38 

effect of altitude, 39 
squalls, 41 

standard weather observation, 80 
surface, 149-150 
thermal wind component, 255 
three forces affecting, 147-148 
trade, 165-166 
triple register, 32 
in the tropics, 208-210 
vanes, 32 

exposure of, 39 
velocity, 34-38 

effect of altitude, 39 
wind-direction indicators, 34 
Wolff, P. M., 259n 

Zones : 

circulation, 161-162, 257 

climatic, 322-325 
equatorial zone, 322 
intermediate, 324 
polar, 324 

subtropical, 323-324 
tropical, 322-325